(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Biological Investigations in the Guadalupe Mountains National Park, Texas"

Biological Investigations 
in the 



Guadalupe Mountains 



National Park, Texas 



Clemson Universi 



3 1604 019 704 123 *T$ '^ 
utrusnuKr l TIM 4 







<&: 






*£&. 






¥£tiU 




••,X' 



36 



l 



t. ,- . ♦ - 




Biological Investigations 
in the 

Guadalupe Mountains 

National Park, Texas 



Proceedings of a Symposium held at Texas Tech University, 
Lubbock, Texas • April 4-5, 1975 



Edited by 

Hugh H. Genoways 
Robert J. Baker 



National Park Service 

Proceedings and Transactions Series Number Four • HI7!) 



For sale by the Superintendent of Documents, U.S. Government Printing Office 
Washington, !>.< '. 20402 



Stock Number 0-M-O05-00744-O 



As the Nation's principal conservation agency, the Department of the 
Interior has responsibility for most of our nationally owned public lands and 
natural resources. This includes fostering the wisest use of our land and 
water resources, protecting our fish and wildlife, preserving the environ- 
mental and cultural values of our national parks and historical places, and 
providing for the enjoyment of life through outdoor recreation. The Depart- 
ment assesses our energy and mineral resources and works to assure that 
their development is in the best interests of all our people. The Department 
also has a major responsibility for American Indian reservation commu- 
nities and for people who live in Island Territories under U.S. admin- 
istration. 



Library of Congress Cataloging in Publication Data 

Main entry under title: 

Biologieal investigations in the Guadalupe Mountains 
National Park, Texas. 

(Proeeedings and transaetions series — National Park 
Serviee ; no. 4) 

Bibliography: p. 

1. Natural history — Texas — Guadalupe Mountains 
National Park. 2. Natural history — Guadalupe Moun- 
tains, N.M. and Tex. 3. Guadalupe Mountains National 
Park. 4. Guadalupe Mountains, N.M. and Tex. 
I. Genoways, Hugh H. II. Baker, Robert J. 111. Series: 
United States. National Park Service. Proeeedings and 
transactions series — National Park Service ; no. 4. 
QH105.T4B56 574.9764 94 78-6061 38 



Contributors 



Gary M. Ahlstrand 

Carlsbad Caverns and 

Guadalupe Mountains National Park 

National Park Service 

Carlsbad, New Mexico 

J. Scott Altenbach 

Department of Biology 
University of New Mexico 
Albuquerque, New Mexico 

John S. Applegarth 

P.O. Box 91 
Durango, Colorado 

Peter V. August 

Department of Biological Sciences 

Texas Tech University 

Lubbock, Texas 

Robert J. Baker 

Department of Biological Sciences 

and The Museum, 

Texas Tech University 

Lubbock, Texas 

Craig C. Black 

The Museum 

Texas Tech University 

Lubbock, Texas 1 

John P. Brand 

Department of Geosciences 

and The Museum 

Texas Tech University 

Lubbock, Texas 



Tony L. Burgess 

Department of Biological Sciences 

Texas Tech University 

Lubbock, Texas 

John W. Clarke 

Department of Biological Sciences 

Texas Tech University 

Lubbock, Texas 

Derrick C. Cooke 

U.S. Forest Service 
Espanola, New Mexico 

John E. Comely 

Department of Biological Sciences 

Texas Tech University 

Lubbock, Texas 2 

Stanley L. Evans 

Carlsbad Senior High School 

103 West Hagerman Street 

Carlsbad, New Mexico 

David E. Foster 

Entomology Department and 

The Museum 

Texas Tech University 

Lubbock, Texas 

Richard W. Fullington 

Dallas Museum of Natural History 

Fair Park Station 

Dallas, Texas 



'Present address: Carnegie Museum of 
Natural History, 4400 Forbes Avenue, Pitts- 
burgh, Pennsylvania 



2 Present address: Department of Biological 

Sciences. Northern Arizona University. Flag- 
staff, Arizona 



in 



Frederick R. Gehlbach 

Department of Biology and 

Institute of Environmental Studies 

Baylor University 

Waco, Texas 

Kenneth N. Geluso 

Department of Biology 
University of New Mexico 
Albuquerque, New Mexico 

Hugh H. Genoways 

The Museum 

Texas Tech University 

Lubbock, Texas* 

Don Harrington 

Dallas Museum of Natural History 

Fair Park Station 

Dallas, Texas 

Alonzo D. Jacka 

Department of Geosciences 

Texas Tech University 

Lubbock, Texas 

Marshall C. Johnston 

Department of Botany 

The University of Texas 

Austin, Texas 

J. Knox Jones, Jr. 

Vice President for Research 

and Graduate Studies 

Texas Tech University 

Lubbock, Texas 

Walter H. Kittams 

1700 West Seventh 
Weiser, Idaho 



Owen T. Lind 

Department of Biology and 

Institute of Environmental Studies 

Baylor University 

Waco, Texas 

Lloyd E. Logan 

The Museum 

Texas Tech University 

Lubbock, Texas 4 

Ernest L. Lundelius, Jr. 

Department of Geology 

The University of Texas 

Austin, Texas 

Paul S. Martin 

Department of Geosciences 

The University of Arizona 

Tucson, Arizona 

M. Houston McGaugh 

Museum Science Program 

The Museum 

Texas Tech University 

Lubbock, Texas 

John S. Mecham 

Department of Biological Sciences 

and The Museum 

Texas Tech University 

Lubbock, Texas 

James V. Moody 

Entomology Department 

Texas Tech University 

Lubbock, Texas 

George A. Newman 

Department of Biology 

Hurdin- Simmons University 

Abilene, Texas 



^Present address: Section of Mammals, 4 Present address: Division of Paleo- 
Carnegie Museum of Natural History, 4400 biology. National Museum of Natural His- 
Forbes Ave., Pittsburgh, Pennsylvania tory, Washington, D.C. 



i\ 



David K. Northington 

Department of Biological Sciences 

and The Museum 

Texas Tech University 

Lubbock, Texas 

Margaret A. O'Connell 

Department of Biological Sciences 

Texas Tech University 

Lubbock, Texas 

Robert L. Packard 

Department of Biological Sciences 

and The Museum 

Texas Tech University 

Lubbock, Texas 

Arthur M. Phillips, III 

Department of Geosciences 

The University of Arizona 

Tucson, Arizona 5 

W. Geoffrey Spaulding 

Department of Geosciences 

The University of Arizona 

Tucson, Arizona 



Thomas R. Van Devender 

Department of Geosciences 

The University of Arizona 

Tucson, Arizona 

Roland H. Wauer 

National Park Service 

Southwest Region 
Santa Fe, New Mexico 

Dallas E. Wilhelm, Jr. 

The Museum 
Texas Tech University 

Lubbock, Texas 

and Science Division 

Lincoln Memorial University 

Harrogate, Tennessee 

Don E. Wilson 

National Bird 

and Mammal Laboratories 

National Museum 

of Natural History 

Washington, D.C. 



spresent address: Biology Department. 
Museum of Northern Arizona. Flagstaff, 
Arizona 



Digitized by the Internet Archive 
in 2013 



http://archive.org/details/biologicalinvestOOtexas 



Editors' Note 

The Guadalupe Mountains National Park, which is located in Trans- 
Pecos Texas between Carlsbad, New Mexico, and El Paso, Texas, is one of 
the newest national parks being formed by the Congress in 1967. The 
Guadalupe Mountains and the associated Chihuahuan Desert included in 
the park represent a unique biological area in which a fragile biological equi- 
librium exists between the fauna and flora of the Chihuahuan Desert of the 
lowlands and the Rocky Mountains of the high elevations. The preservation 
of this area will depend upon sound management decisions. 

This volume is the result of a symposium held at The Museum of Texas 
Tech University on 4 and 5 April 1975. The impetus for this symposium was 
furnished by our participation in a research project in the Guadalupe Moun- 
tains National Park. This work was funded by the Southwest Region of the 
National Park Service through Texas Tech University and administered by 
Mr. Roland H. Wauer. It became apparent to us that considerable scientific 
research was being conducted in the Park both by the staff of Texas Tech 
University and at other institutions. However, we believed that only limited 
communication was occurring between these scientists. We hope that the 
symposium and this volume will significantly enhance this scientific com- 
munication and will provide much of the baseline data necessary for the 
development of a master plan by the National Park Service for the Guada- 
lupe Mountains National Park, Texas. Because the Carlsbad Caverns 
National Park, New Mexico, is geographically and physiographically 
closely related to the Guadalupe Mountains, we have included several 
studies that recently have been conducted there. 

The research resulting in this symposium would not have been possible 
without the cooperation of the National Park Service. Mr. Roland H. 
Wauer, Regional Chief Scientist, Southwest Region, has been instrumental 
in initiating many of these research projects, especially those being con- 
ducted by personnel of Texas Tech University. Special thanks are due the 
personnel of the Guadalupe Mountains and Carlsbad Caverns National 
Parks, especially Donald A. Dayton, Superintendent; John Chapman, Area 
Manager; Gary M. Ahlstrand, Research Ecologist; Philip F. Van Cleave, 
Staff Interpretative and Environmental Services Specialist; and Roger 
Reisch, Park Ranger. 

vii 



The symposium at Texas Tech University was sponsored by The Museum, 
the Graduate School, and the International Center for Arid and Semi-arid 
Lands Studies. We wish to express our gratitude to Dr. Craig C. Black, Dean 
J. Knox Jones, Jr., and Dr. Frank B. Conselman, respectively, of these 
organizations. The editors wish to extend their personal thanks to Dr. Craig 
C. Black, not only for his support of the symposium, but also for his support 
of our activities throughout our work in the Guadalupe Mountains National 
Park. We also gratefully acknowledge the editorial assistance and attention 
to detail of R. Laurie Hendricksen. Stephen L. Williams aided with some of 
the illustrative material. Many of the proper names used for specific locali- 
ties are defined in the paper on mammals, whereas others may be found on 
the U.S. Geological Survey quadrangle map for Guadalupe Peak, Texas 
(1:62,500, 1933). 

Hugh H. Genoways 
Robert J. Baker 



Vlll 



Welcome 



It gives me great pleasure to welcome participants and guests to Texas 
Tech University for this important symposium. The Guadalupe Mountains 
National Park provides a unique outdoor laboratory for biological investi- 
gations, which, of course, will provide the baseline data for the National 
Park Service in developing an interpretive program for public use of this 
area. Because investigators at Texas Tech University have played a major 
role in conducting biological studies in the park, it is especially appropriate 
that this institution take the lead in promoting an interchange of informa- 
tion for the benefit of the scientific community and also the more pragmatic 
interests of the National Park Service. 

In the past decade or so, we have experienced a trend in the biological 
sciences in this country toward concentration of investigative research in the 
several experimental fields. At the same time, there has been a tremendous 
increase in interest on the part of the general public in environmental quality, 
preservation of natural areas, and in public use of national parks, monu- 
ments, and other similar facilities. It occurs to me that it is paradoxical, given 
a clear "back to nature" movement within our society today, that few insti- 
tutions of higher learning are emphasizing the training of environ- 
mentalists. I am pleased to note that at Texas Tech University there still is a 
strong tradition in field-oriented studies by professors and students in our 
Department of Biological Sciences, in several departments in the College of 
Agricultural Sciences, and in our Museum Science program. The location of 
a large, new museum at the University contributes materially to this thrust. 

I congratulate the organizers of this symposium and express my gratitude 
to those of you who are participants and attendees for supporting their 
efforts. In perusal of the program, I note not only that the symposium will 
bring together a substantial body of information on one of our newest 
national parks but also that the subject matter underscores the significant, 
continuing, and pressing need for research and teaching at the university 
level in ecology, natural history, and systematics of plants and animals, and 
in related areas of the geosciences and anthropology. 

I wish for you a most successful and productive conference. 

J. Knox Jonls, Jr. 
ix 



Introduction 



The reality of this Symposium on the Biological Investigations in the 
Guadalupe Mountains National Park gives me a great deal of personal satis- 
faction. I am proud to be part of the program. 

Guadalupe Mountains have been a national park only since 1966, and 
before that time the area received attention from only a few scientists. The 
greatest amount of study resulted from private exploration for oil and gas. 
Vernon Bailey's 1905 report represents the earliest biological survey of the 
Guadalupe Mountains. Barton Warnock spent considerable time studying 
the plant life of the area, particularly within the lower canyons. William 
Davis studied the mammals of the Guadalupes. Fred Gehlbach's investiga- 
tions of the herpetofauna was the first ecological analysis of the region. 
Owen Lind's limnological studies in McKittrick Canyon are continuing. 

Since 1966, the Guadalupe Mountains have become a scientist's beehive. 
Titles of papers to be presented at this symposium demonstrate that research 
in the area is as varied as the Guadalupe environments. 

We in the National Park Service believe that the recent series of investiga- 
tions initiated within the Guadalupes are the proper way to develop a sound 
research and resources management program for this park, and to provide 
resources information for park planners, interpreters, and managers. 

Too often Federal bureaucrats expend energy and great sums of money 
planning park developments without due regard to the full protection of the 
area's resources that were the primary reasons for establishment of the park. 
Too often land management agencies place fragile and unique resources 
second to roadways, campgrounds, visitor centers, and other facilities. Too 
seldom do we attempt to understand fully the potential implications of a 
development within a natural system. 

Our approach at Guadalupe Mountains National Park began with the 
development of research priorities. This list is included within a Natural 
Resources Management Plan that is revised annually. Copies of the 1975 
Guadalupe Mountains National Park Natural Resources Management Plan 
are available. The plan, initiated by the park staff and coordinated by the 
Office of the Regional Chief Scientist, includes two major sections: 

xi 



1. Project statements which identify the area's natural resource prob- 
lems and requirements to comply with National Park Service natural 
science standards. 

2. Five-year programs that include maintenance costs, man-year needs, 
research priorities, and cost estimates. 

A Natural Resources Management Plan is prepared for every National Park 
Service area containing natural resources. 

Research priorities for Guadalupe Mountains National Park include the 
following items: 

1. Inventory of Flora; 

2. Fire Ecology Study; 

3. Inventory of Fauna; 

4. Climatological Data; 

5. Inventory of Significant Geological Features; 

6. Vegetative Analysis; 

7. Faunal Factors; 

8. Data Analysis; 

9. Human Intrusion on the Ecosystem; 

10. Soils Inventory and Analysis; 

11. Water Resources Analysis; 

12. Inventory of Microorganisms; and 

13. Ecosystem Analysis. 

This list includes only those studies to be completed and new projects 
required to comply with natural science standards. A complete ecosystem 
analysis of the Guadalupe Mountains may take decades to complete. We do 
plan to reach the ecosystem analysis stage at a few of our smaller natural 
parks as other project priorities are completed. Southwestern areas such as 
Bandelier, Capulin Mountain, and White Sands may reach that stage of 
comprehension fairly soon. 

The Guadalupe Mountains plan calls for an annual funding base of at 
least $25,000 for research. Admittedly, the funding of $25,000 annually to 
field a team of scientists to gather information on an area as diverse as the 
Guadalupe Mountains is not very much. However, the National Park 
Service will continue to fund for many years the work that Texas Tech and 
its associated scientists have started in the Guadalupe Mountains. We 
believe that our association with University-based scientists is the best pos- 
sible route to the completion of a basic resources inventory, analysis of those 
resources, and the development of a management information system. 

The first objective is to acquire a resources basic inventory. We are espe- 
cially interested in those resources that are unique or of special visitor 
interest. We will hear about some of the floral and faunal resources in the 
papers to follow. It is imperative to establish environmental baselines for use 
in long-range monitoring. The vegetative transects and mammalian grids 
installed within the Guadalupe Mountains system will function as a future 
xii 



warning system against environmental impacts. The quality of the baseline 
studies must be high. 

There are cases when special emphasis on project continuity and coordi- 
nation of data requires other than seasonally oriented studies. We believe 
that a resident research scientist is a necessary element of the Guadalupe 
Mountains National Park staff. Dr. Gary Ahlstrand is investigating fire 
management possibilities within the Guadalupes; he also advises the Park 
Superintendent on research priorities, evaluates project results, analyzes 
their implications in the potential role of fire in the ecosystem, and is 
developing a fire-management program based upon his research and related 
data. Fire in the Guadalupe Mountains ecosystem probably will command 
an important position in management guidelines for the retention of that 
natural system. 

Human intrusion on the Guadalupe Mountains is another concern that is 
being addressed within all of the studies. Carrying-capacity determinations 
will be made on sound biological information that is obtained by scientists 
concentrating on their various disciplines. The assimilation of those data 
with park requirements will lead toward decisions that will cause minimal 
impact. 

Computerization of all the data obtained through field studies and the 
laboratory began during the first funded year. A coordinate grid system of 
31,377 cells was developed for the 77,500-acre park. The purpose of struc- 
turing a data bank is to describe each data cell as completely as possible in 
terms of physical, cultural, and natural resources. Data storage and retrieval 
must be compatible with the National Park Service computer program 
housed at NASA's Slidell Computer System in Mississippi. 

The National Park Service Science Center near Slidell is responsible for 
developing a management information system. This involves integrating 
scientific expertise, archival materials, and computer sciences with manage- 
ment and planning activities. It is our intention that, within the decade, 
Washington, D.C., and regional offices and major parks will possess ter- 
minals so that park staffs can take advantage of activities currently under- 
way to manage soundly and to interpret their areas. 

An initial Management Information System is already in use for back- 
country campground management at Great Smokies National Park. A simi- 
lar system has been installed at Grand Canyon for river management. Time 
and money will be needed for all parks to join such a system. The Guadalupe 
Mountains National Park program is a beginning. 

ROLAND H. WAl ER 



Kill 



Contents 



Editor's Note vii 

Welcome — J. Knox Jones, Jr ix 

Introduction— Roland H. Wauer xi 

Geology of the Guadalupe Mountains National Park 

John P. Brand and Alonzo D. Jacka 1 

Late Pleistocene Plant Communities in the 
Guadalupe Mountains, Culberson County, Texas 
Thomas R. Van Devender, W. Geoffrey Spaulding, and 
Arthur M. Phillips, III 13 

Preliminary Report of the Ecology of Fire Study, 
Guadalupe Mountains and Carlsbad Caverns National 
Parks 
Gary M. Ahlstrand 31 

The Guadalupe Mountains— A Chink in the Mosaic of the 
Chihuahuan Desert? 
Marshall C. Johnston 45 

Summary of the Vegetative Zones of the Gua-dalupe 
Mountains National Park, Texas 
David K. Northington and Ton y L. Burgess 51 

Status of Rare and Endangered Plant Species of the 
Guadalupe Mountains National Park/Texas 
David K. Northington and Tony L. Burgess 59 

Agave— Complex of the Guadalupe Mountains National 
Park: Putative Hybridization Between Members of 
Different Subgenera 
Tony L. Burgess 79 

The Land and Freshwater Mollusca of the Guadalupe 
Mountains National Park, Texas 
Richard W. Fullington 91 

Plusiotis woodi AND Plusiotis gloriosa (SCARABAEIDAE): 
First Report of the Guadalupe Mountains 
National Park 
Richard W. Fullington and Don Harrington 113 



xv 



Notes on the Bionomics and Nest Structure of 
Pogonomyrmex maricopa HYMENOPTERA: FORMICIDAE) 
James V. Moody and David E. Foster 115 

Limnology of McKittrick Creek, Guadalupe 
Mountains National Park, Texas. 
Owen T. Lind 1 23 

The Quaternary Vertebrate Fauna of Upper Sloth Cave, 
Guadalupe Mountains National Park, Texas 
Lloyd E. Logan and Craig C. Black 141 

Environmental Implications of Herpetofaunal Remains 
from archeological sites west of carlsbad, 
New Mexico 
John S. Applegarth 159 

The Biogeographical Relationships of the Amphibians and 
Reptiles of the Guadalupe Mountains 
John S. Meeham 1 69 

Compositional Aspects of Breeding Avifaunas in 
Selected Woodlands of the Southern Guadalupe 
Mountains, Texas 
George A. Newman 181 

Post-Pleistocene Mammals from Pratt Cave and 
Their Environmental Significance 
Ernest L. Lundelius, Jr 239 

Ground Sloth Dung of the Guadalupe Mountains 

W. Geoffrey Spaulding and Paul S. Martin 259 

Mammals of the Guadalupe Mountains National Park, 
Texas 

Hugh H Genoways, Robert J. Baker, and 
John E. Comely 271 

Demographic Patterns of Small Mammals: A 
Possible Use in Impact Assessment 
Peter V. August, John W. Clarke, M. Houston 
McGaugh, and Robert L. Packard 333 

Population Size of Tadarida brasiliensis at 
Carlsbad Caverns in 1973 
J. Scott Altenbach,, Kenneth N. Geluso, and Don E. Wilson ..341 



xvi 



Coexistence of Two Species of Kangaroo Rats (Genus 
Dipodomys) in the Guadalupe Mountains National 
Park, Texas 
Margaret A. O'Connell 349 

Ecological Distribution of Woodrats (Genus Neotoma) 
in Guadalupe Mountains National Park, Texas 
John E. Comely 373 

Status of the Guadalupe Mountains Vole, Microtus 
mexicanus guadalupensis 
Dallas E. Wilhelm, Jr 395 

Food Habits of Mule Deer on Foothills of Carlsbad 
Caverns National Park 
Walter H. Kit tarns, Stanley L. Evans, and Derrick C. Cooke . . . 403 

blomes of the guadalupe escarpment: vegetation, 
Lizards, and Human Impact 
Frederick R. Gehlbach 427 

Research in National Parks 

Robert J. Baker and Hugh H. Genoways 441 



xvn 



Geology of the Guadalupe Mountains 
National Park 



JOHN P. BRAND and ALONZO D. JACKA 

Texas Tech University, Lubbock 

The Guadalupe Mountains National Park enjoys a worldwide reputation 
as a region in which a unique complex of geological phenomena may be 
viewed and studied. Rocks, ranging in age from possibly medial Precam- 
brian to Holocene crop out in or within the near vicinity of the park. Geolo- 
gists have pieced together a fairly complete history of the last one and one- 
half billion years through study of these rocks. Although the region has been 
viewed by geologists for over 100 years, detailed studies did not begin until 
about 50 years ago; the renewed interest was prompted by the exploration 
for and production of petroleum and natural gas in west Texas. 

The Guadalupe Mountains National Park is situated within an area com- 
monly assigned to the Trans-Pecos region of Texas. This, in turn, occupies a 
portion of a well-known geological province, the Permian Basin. Trans- 
Pecos Texas is a land of contrasts, both physiographically and geologically. 
The region contains numerous distinct mountain ranges and intervening 
basins (Fig. 1). Some ranges are folded and block faulted, the product of the 
late Cretaceous to early Cenozoic Laramide Revolution and middle-to-late 
Cenozoic block faulting. Accordingly, the region possesses affinities with the 
Rocky Mountains and the Basin and Range Province of western America. 
Several ranges are made up of thick accumulations of lava flows and pyro- 
clastic ejecta. Finally, the region boasts remnants of the folded Marathon 
Mountains, the product of late Paleozoic orogeny in the Ouachita- 
Marathon geosynclinal belt. Elements of the structure and stratigraphy of all 
the above are involved in the reconstruction of the history of Guadalupe 
Mountains National Park. To turn attention to the region of the park, one 
first must view the area of the expanded Permian Basin (Fig. 2). This maze of 
paleotectonic features was in the process of formation during most of the 
Paleozoic Era, but the features did not assume their Permian form until 
during late Pennsylvanian and early Permian time. Orogeny and sub- 
sequent erosion during this latter interval produced numerous platforms, 
shelves, and basins which controlled the types of sedimentary rocks that 

1 



RAND AND JACK. A 



E>- 




Fis. 1. Ranges and basins of part of Trans-Pecos Texas. After Permian Basin 
Section. S. E. P. M. 1975. 



accumulated during the Permian Period (Fig. 3). With inundation of the 
complex topography by the early Permian sea evolved the possibility of 
deposition of sediments in such diverse environments as shallow-water plat- 
forms, back-reef lagoons, and deep-water basins. The margins of the 
shallow-water platforms afforded an opportunity for reef-forming 
organisms to flourish. Subsurface exploration has established the presence 



GEOLOGY 




OKLAHO MA \ V 

frFfW"] - | .ooow.no | • 

««««! HANSfOWO ' OCMIITHCII HP»OOM» . I 



VEGAS I ■ tf - 
BASIN I " Amomt 



■ •OIL ,<*■ V ,' MS I 




j ««»T.|H^Ha L f * "' j i 

I I IKOOCK CUITCR I 

n — i 



»LL I »RMJT«0nH °OW.tT > U«J. 1 l«»- | 1 I 

I 5' I — ■ — ' \| | '»""' 'p»*MCR I C4STW0 I SWISHER I BRISCOC I HAH. IcHILOWtSS \^-^/ V J 




Fig. 2. Paleotectonic features of Permian Basin. 



of a barrier reef along the margins of the Diablo Platform, the Northwest 
Shelf, the Central Basin Platform, and the Eastern Shelf. The entire North 
American continent records no barrier reefs of similar magnitude. Behind 
the reefs were wide, shallow lagoons in which carbonate, evaporite, and 
clastic sediments accumulated. Immediately seaward of the reefs were basins 
up to 2000 ft deep in which accumulated a thick sequence of predominantly 
clastic sediments which were transported down to the basin through sub- 
marine canyons and deposited as deep sea fans. On the slope between the 
reefs and the basin, accumulated piles of talus derived from the seaward face 



BRAND AND JACKA 




Outcropt of Wolfcomp, 
Leonard, Guadalupe, 
and Ochoa Series 
(Permian), ond their 
equivalents 



Boundaries of prov- 
inces of Permian 

time 



asln areas 



Deformed pre-Wolfcomp 
Paleozoic rocks of 
Marathon folded belt 



Fig. 3. Permian geological provinces of west Texas and southeast New Mexico. 



of the reefs (Fig. 4). The back-reef lagoon, the reef, and the fore-reef talus 
constitute a stratigraphically restricted but complex relationship referred to 
as the "Permian Reef Complex" (Fig. 5). 

The visitor to the Guadalupe Mountains National Park may view all the 
aforementioned environments of deposition via a leisurely walk through any 
of the major canyons which trend normal to the Guadalupe escarpment. For 
the motorist driving from Carlsbad, New Mexico, to Guadalupe Pass, the 
following applies. From Carlsbad to White City, one drives quite near the 
margin of the reef facies. From White City to Guadalupe Pass one travels 



GEOLOGY 5 







Fig. 4. El Capitan. 




Basin facies 



Fig. 5. Diagrammatic sketch of Permian Reef Complex. 



over rocks which were deposited in the Delaware Basin. The most conspicu- 
ous unit, the Castile Gypsum, crops out in a road cut where the highway 
descends the Yeso (Spanish word for gypsum) hills to the flat area a mile 
north of the New Mexico-Texas state line. Farther east, in the subsurface, 
one would encounter extensive beds of halite (rock salt) with which are asso- 
ciated the potash minerals of the Carlsbad mining district. Both the gypsum 
and halite were formed as the result of evaporation of sea water. West of the 
highway, one cannot but be impressed by the precipitous Guadalupe escarp- 
ment. The top of the escarpment is formed by the Capitan barrier reef; the 
beds which slope steeply toward the valley floor are within the fore-reef talus 



BRAND AND JACKA 



. 



^ 




Fig. 6. West front of Delaware Mountains. 



facies. The nearly horizontal beds near the highway are rocks of the basin 
facies. The rock in El Capitan was deposited near the margin of the reef 
barrier and the fore-reef talus. A glance to the southeast from Guadalupe 
Pass affords a view of the Delaware Mountains— a range made up of clastic 
rocks of the Delaware Basin facies (Fig. 6). 

Specific aspects of the geology of the Guadalupe Mountains involve two 
separate studies. The first relates to the sedimentological history of the Per- 
mian Period as interpreted from the rocks in the immediate area. The 
second, mainly physiographic, concerns the late Tertiary and Quaternary 
history of the region. Following are brief discussions of the salient features of 
these two rather widely separated intervals of geologic time. 

PERMIAN HISTORY 

The earliest events of the Permian Period are recorded in rocks of lower 
Permian (Wolfcampian) age. The first unit, the Hueco Formation truncates 
rocks of nearly all ages from Precambrian through late Pennsylvanian. The 
basal member, the Powwow Conglomerate, contains debris from nearly all 
pre-Permian units. Wolfcampian beds contain predominantly carbonate 
units across the platforms and clastic units in the basins. They do not, how- 
ever, display the distinct differentiation into reef complexes evident in later 
Permian strata. One overriding characteristic of the Permian sea was the ten- 



GEOLOGY 



SHELF 



MARGIN 




DEWEY LAKE REDBEDS 
RUSTLER FORMATION 
SALADO FORMATION 



TANSILLFM. 



YATES FM. 



SEVEN 

RIVERS 

FM. 



QUEEN FM. 



GRAYBURG 
FM. 



CAPITAN 
LIMESTONE 



GOAT SEEP 
LIMESTONE, 




BASIN 



DEWEY LAKE REDBEDS 
RUSTLER FORMATION 
SALADO FORMATION 
CASTILE FORMATION 



Lamor Ls Mbr. 



MCCombs Ls. Mbr 



BELL 



Manzanita Ls. Mbr. 



•CHERRY CANYON 



South Wells Ls. Mbr 



m$ FORMATION^ 



Getaway Ls. Mbr. 



;//>:': Sandstone '. Tongue :pf . Cherry .Canyon, Fm. : ;., , ; 



SAN ANDRES 
LIMESTONE 



:; >BRUSHY CANYON 
:■;■:}: FORMATION ■}■;■: 



■. Pipeline • Shale. Mbr. 



CUTOFF SHALE 



VICTORIO PEAK 
LIMESTONE 



BONE SPRING 
LIMESTONE 



Fig. 7. Shelf, margin, and basin facies of Permian Basin. 



dency, throughout Permian time, for the sea in the basin to become progres- 
sively smaller and deeper. The regression of the sea was not at a uniform rate, 
as many transgressive and regressive oscillations are recorded. Some of these 
sea-level fluctuations were undoubtedly of a eustatic nature and appear to 



5 BRAND AND JACKA 

reflect sea-level changes because of multiple glacial cycles during the 
Permian Period. 

At the beginning of the second Permian Epoch, the Leonardian, the con- 
trast of shallow-water platform and deep-water basin topographies was well 
established and facies differentiation so characteristic of Leonardian and 
Guadalupian units became evident (Fig. 7). The first barrier reef of signifi- 
cant size, the Victorio Peak reef, developed along the eastern margin of the 
Diablo Platform. During growth of the Victorio Peak reef, a dark-gray to 
black limestone (Bone Springs Limestone) accumulated in the deep water of 
the Delaware Basin. This petroliferous unit may be seen in the first road cut 
as one approaches Guadalupe Pass from the south. Simultaneously, a 
lagoonal-arid coastal plain facies of gypsum, dolomite, and siltstone of the 




HOVEY 
CHANNEL 



Fig. 8. Submarine canyons and deep-sea fans of Permian Basin. 



GEOLOGY 9 

Yeso Formation accumulated behind the Victorio Peak reef. The axis of the 
Victorio Peak reef lies several miles west of the axis of the first Guadalupian 
reef. Although outcrops of the Victorio Peak may be seen along the western 
side of the Guadalupe Mountains in Shirttail Canyon and Shumard 
Canyon, the best exposures are in Victorio Canyon in the Diablo Plateau 
about midway between Guadalupe Pass and Van Horn, Texas. Two distinct 
stages of reef growth are recorded during the Guadalupian Epoch — the first, 
the Goat Seep reef; the second, the Capitan reef. The Goat Seep reef lies sev- 
eral miles west of the Capitan reef. The spatial relationship between axes of 
the Victorio Peak, the Goat Seep, and the Capitan reefs attests to the gen- 
eral shrinking of the Permian sea and the consequent regressive (offlap) rela- 
tionship of these Permian reefs. Because of its accessibility and unusually 
clean exposures, the Capitan reef has been studied in considerably more 
detail than other Permian reefs. The major canyons — McKittrick, 
Slaughter, Dark, Walnut, etc. — trend nearly at right angles to the reef axis. 
Detailed studies have revealed that the typical Permian reef consists of two 
basic fractions. The first is the framework builders — organic remains which 
afforded rigidity and resistance to wave action; the other is sediment which 
became trapped by the framework. The formation might be likened to 
tumbleweeds and other vegetation trapping wind-blown sand in the present 
West Texas environment. Reef-forming organisms consisted of such diverse 
forms as various algae, corals, bryozoans, brachiopods, and sponges. Algae 
and calcareous sponges were probably the greatest contributors to the reefs. 
The reefs, although elongate in overall form, were not continuous barrier- 
like ramparts. Their contours were interrupted by deep transverse channels 
which extended reefward from the deep water of the Delaware Basin 
through the reef and into the back-reef lagoon. These submarine canyons, 
believed to be essentially co-linear with the present transverse valleys, 
became avenues for transportation of clastic materials from the back-reef 
lagoons into the Delaware Basin. Units of the Delaware Mountain Group, 
the Brushy Canyon, Cherry Canyon, and Bell Canyon contain sands and 
shales transported by "turbidity currents" into water as deep as 2000 ft. The 
basin clastic units are in reality subsea fans (Fig. 8) which possess primary 
sedimentary structures identical to those observed in modern deep-sea fan 
accumulations that extend beyond continental slopes. Eustatic fluctuations 
of sea level are recorded in the Delaware Basin by alteration of thick clastic 
intervals (glacial maxima) and thin limestone layers (interglacials) which 
extend miles into the basin. The uppermost unit exposed in the walls of the 
large canyons, the Tansill limestone, records a progradation of back-reef 
limestone eastward across the top of the reef. This, again, is another indica- 
tion of the general regressive tendency of Permian reefs. 

At the end of the Guadalupian Epoch, one might visualize the Guadalupe 
Mountains National Park as follows. The shoreline was nearly coincident 
with the crest of the present Guadalupe escarpment. The back-reef lagoon to 
the west was mainly emergent. To the east, the Delaware Basin remained 



10 BRAND AND JACKA 

unfilled by sediments, but contained water over 1000 ft deep. The basic "reef 
complex" physiography was still present. 

At the beginning of the Ochoan Epoch, the Permian Basin appears to have 
tilted upward in the west and downward to the east, along an axis in the 
eastern part of the Delaware Basin. As a result, the Guadalupe Mountains 
National Park area became tilted upward and may have been emergent, 
whereas the eastern part including the Central Basin Platform became more 
deeply submerged. This "dead sea" received only periodic replenishment of 
water and the extremely arid climate of the time created a rate of evapora- 
tion considerably in excess of the rate of influx of water of normal sea-water 
salinity. With continued evaporation and eventual supersaturation of 
calcium sulfate came deposition of the Castile anhydrite and gypsum. Con- 
tinued evaporation led to deposition of the Salado Salt and associated 
potash minerals. A final effort of transgression by the sea afforded deposi- 
tion of the Rustler Formation. Concommitant with the withdrawal of the 
Rustler Sea occurred the deposition of the fluviatile sediments of the Dewey 
Lake Formation. If one might have viewed the Permian Basin at the end of 
the Ochoan Epoch, he would have seen no high Guadalupe Mountains. The 
Delaware Basin was filled, the Central Basin Platform was covered, the Mid- 
land Basin was filled, and the Capitan reef was buried. The Guadalupe 
Mountains were no more than low knobs which rose only slightly above a 
broad, flat depositional plain. 

MESOZOIC HISTORY 

Triassic, Jurassic, and Cretaceous history adds little to the story of the 
Guadalupe Mountains National Park except a late Triassic subsidence of the 
Delaware Basin, which permitted the deposition of a rather thick sequence 
of the fluviatile deposits of the late Triassic Dockum Group. The area 
probably was completely emergent during the Jurassic Period. Cretaceous 
rocks were deposited over the area, but subsequent erosion has removed all 
save a few remnants in the Delaware Basin and across the Central Basin Plat- 
form. Thicker and more continuous sequences of Cretaceous rocks may be 
seen in areas south of the Guadalupe Mountains. 

CENOZOIC HISTORY 

The Permian Basin along with the balance of Trans-Pecos Texas 
remained emergent following the close of the Cretaceous Period. In some 
places, notably the Guadalupe Mountains and the Llano Estacado, east- 
ward flowing streams removed nearly all vestiges of Cretaceous rocks. To 
the south, volcanic activity produced the huge volcanic piles now seen in the 
Davis Mountains, the Sierra Vieja, and in the Big Bend region. During late 
Miocene and early Pliocene time, western Texas, along with most of western 
North America, was subjected to nearly vertical uplift and accompanying 
block mountain and graben basin faulting. The spectacular topography of 
the west side of the Guadalupe Mountains as well as that along the front of 



GEOLOGY 1 1 

the Diablo Plateau at the margin of Salt Basin were produced by this 
diastrophic episode. The eastern Guadalupes and the balance of the Permian 
Basin were not faulted, but uplift permitted groundwater to permeate and 
dissolve much of the Salado and Castile formations in the Delaware Basin. 
With removal, by solution, of the evaporite sequence, collapse of the over- 
lying Permian, Triassic, and Cretaceous rocks produced a huge closed 
depression commonly called the "Pecos Depression." Debris eroded from 
the mountainous terrain to the west accumulated in and ultimately filled the 
"Pecos Depression" to overflowing and the blanket of fluviatile sand, gravel, 
and clay extended eastward to form the Ogallala Formation of the Llano 
Estacado. Preliminary evidence indicates that the surface of the High Plains 
was graded initially to the tops of the highest peaks in the Guadalupe Moun- 
tains. Again, as at the close of the Permian Period, the Guadalupe Moun- 
tains were buried, this time by Pliocene sands and gravels. Near the 
beginning of the Pleistocene Epoch, headward erosion by the Pecos River 
pirated the "Pecos Depression." With progressive deepening of the Pecos 
Valley and consequent removal of loosely consolidated sediment by stream 
erosion and transportation, the Guadalupe Mountains have been exhumed. 
No references have been cited in this report as it is intended to present only 
a "thumb nail" sketch of salient features of the geological history of the 
Guadalupe Mountains National Park area. For more detailed discussions of 
the structure and stratigraphy of the region, the reader is referred to the 
following. 

SELECTED REFERENCES 

Adams, J. E., and H. N. Frenzel. 1950. Capitan barrier reef, Texas and New 
Mexico, J. Geol, 58:289-312. 

King, P. B. 1948. Geology of the southern Guadalupe Mountains, Texas. U.S. Geol. 
Surv. Prof. Paper, 215:1-183 and maps. 

1959. The Evolution of North America. Princeton Univ. Press, Princeton, 

New Jersey, 190 pp. 

Lloyd, E. R. 1929. Capitan Limestone and associated formations of New Mexico 
and Texas. Am. Assoc. Petrol. Geol. Bull. 13:645-658. 

Newell, N. D., et al. 1953. The Permian Reef Complex of the Guadalupe Moun- 
tains Region, Texas and New Mexico: A Study in Paleoecology. W. H. Freeman 
& Co., San Francisco, 236 pp. 



Late Pleistocene Plant Communities 
in the Guadalupe Mountains, 
Culberson County, Texas 



THOMAS R. VAN DEVENDER, W. GEOFFREY 
SPAULDING, and ARTHUR M. PHILLIPS, III, University 
of Arizona, Tucson 



For half a century the caves along the Guadalupe escarpment in New 
Mexico and Texas (Fig. 1) have provided exciting records of past biotas. 
Rich bone deposits in Burnet Cave (Howard 1932; Schultz and Howard 
1935), Williams Cave (Ayer 1936), Dry Cave (Harris 1970), Pratt Cave 
(Gehlbach and Holman 1974), and the High Sloth Caves (Logan and Black, 
this volume) have documented dramatic changes in the mammalian fauna of 
the Guadalupe Mountains over the past 10,000 years. Large extinct animals 
including horse (Equus sp.), camels (Camelops, Tanupolamd), and Shasta 
ground sloth (Nothrotheriops shastense) have been recovered. Bones of 
several small mammals, now restricted to high-elevation forests to the north, 
including the yellow-bellied marmot (Marmot a flaviventris, Stearns 1942), 
bushy-tailed packrat (Neotoma cinerea, Harris 1970; Logan and Black, this 
volume; Schultz and Howard 1935). and the masked shrew (Sorexcinereus, 
Logan and Black, this volume), also have been recovered from these caves. 
However, most of these vertebrate fossils were collected in moist limestone 
caves where plant macrofossils and pollen are not commonly preserved. 
Because of this lack, most of the present theories on the Ice Age environ- 
ments and paleoecology of the Guadalupe Mountains were based solely on 
the vertebrate record. 

In this paper, we present a chronological sequence of late Pleistocene and 
Holocene plant communities spanning the last 13,000 years. The chronology 
is based upon plant macrofossils and pollen taken from the Upper Sloth 
Caves (C-05, C-08, and C-09, designations given by the National Park 
Service) and from Williams Cave (Fig. 1). These materials have been studied 
with the permission and support of the National Park Service and Guada- 
lupe Mountains National Park. Most of the fossils from the Upper Sloth 
Caves were collected in conjunction with the recent excavation for verte- 

13 



14 



VAN DEVENDER ET AL. 



• "N NEW 
Albuquerque \ MEX |C0 

\ 

\ 

I 




TEXAS 


/ 
( 

1 
\ 
1 
1 




Lubbock 


Sierra B 1 anca i 




• 


A • \ R o s w e 1 1 






C\Sq Mo ,' 






1) rv ! 

V\5C ' 
Gu MoVAa t ^Carlsbad 




N 


usck\ DC \ 




i 


Mi El Paso \*¥*pc \ 

s s?o' /.'^N Davis 

\ \;/ MtS. 

\ 

1 
\ 


V 

\ 
\ 
< 


100 km. \ 
• ' s Chisos 


/ 
/ 


N \ s Mts. 


/ 
/ 


CHIHUAHUA 7"-' 


COAHUILA 



Fig. 1. Map of important localities mentioned in text. GuMo - Guadalupe 
Mountains, SaMo = Sacramento Mountains, BC = Burnet Cave, DC = Dry Cave, PC 
= Pratt Cave, WC = Williams Cave, USC = Upper Sloth Caves, N = north. 



brate foss'ls by Lloyd E. Logan of Texas Tech University. The plant macro- 
fossils will be deposited in the collections of the National Park Service, 



PLEISTOCENE PLANTS 15 

Carlsbad, New Mexico. Plant nomenclature used in this paper follows 
Correll and Johnston (1970); tree distributions follow Little (1971). 

BIOCHRONOLOGICAL ZONES 

The plant communities in the Guadalupe Mountains have gradually 
changed from relatively mesic woodland and forest associations during 
pluvioglacial climates in the late Wisconsin Glacial Epoch to the present 
xeric Chihuahuan desertscrub. This transition can be segregated into two 
late Pleistocene (Wisconsin 1 and 2) and at least two Holocene (Holocene 1 
and 2) biochronological zones characterized by particular plant macro- 
fossils (Tables 1, 2) and pollen assemblages (Fig. 2, Tables 3, 4). The macro- 
fossil and pollen data are placed into a time perspective by stratigraphy and 
by radiocarbon dating of carefully selected material from known strati- 
graphic contexts. A summary of the characteristic plant species, pollen 
assemblages, and radiocarbon dates is presented in Table 2. 

Wisconsin 1 A. — The oldest samples of plant macrofossils and pollen were 
collected from an indurated fossil packrat (genus Neotoma) midden and 
from cave fill in two of the Upper Sloth Caves (C-08, C-09). These caves are 
located at 2000 m elevation below Shumard Peak on the steep west face of 
the Guadalupe Mountains. The macrofossils record a subalpine forest with 
Picea sp. (spruce), Juniperus sp. (juniper), J. communis (dwarf juniper), 
Pseudotsuga menziesii (Douglas fir), Pinus strobiformis (southwestern 
white pine, including P. flexilis of other authors), P. edulis (Colorado 
pinyon), Ostrya knowltonii (hop-hornbeam), Quercus gambelii (Gambel 
oak), Arctostaphylos sp. (manzanita), Robinia neomexicana (New Mexican 
locust), and Rubus strigosus (raspberry). Two radiocarbon dates on Picea 
sp. needles were 13,000 ± 730 (A-l 539) and 13,060 ±280 (A- 1549) radiocar- 
bon years before present (BP). The present vegetation is a complex, 
high-elevation Chihuahuan desertscrub mixed with chaparral and grassland 
species. The only trees presently near the Upper Sloth caves are two relict 
Pinus edulis in protected, shady spots. The pollen assemblages (relative 
percentages) associated with the Zone W1A macrofossils are characterized 
by moderate Picea, moderate to high Ostrya, high Gramineae, and low 
Cheno-ams and short-spine Compositae (Fig. 2). Rubus-type Rosaceaeand 
Ribes pollen are present. 

Picea sp., Juniperus communis, and Rubus strigosus are no longer present 
in the flora of Texas. Picea sp. {P. engelmannii and P. pungens) and Rubus 
strigosus presently reach their southern limit in the Sacramento Mountains, 
New Mexico, 1 10 km to the north (Fig. 3; Little 1971). Picea chihuahuana 
(Chihuahuan spruce) presently occurs near Creel, Chihuahua, Mexico, 450 
km to the southwest in the Sierra Madre Occidental at about 2215 m 
elevation. However, this species is distantly related to the Rocky Mountain 
species and certainly represents a much earlier distributional separation than 
the late Wisconsin (Gordon 1968). The distribution of Rubus strigosus 
differs somewhat in that it is present in pine forest and pine-oak woodland 



16 



VAN DEVENDER ET AL. 



TABLE 1. Late Pleistocene plant macrofossils recovered from cave fill and fossil packrat 
middens in the Guadalupe Mountains, Culberson County, Texas. WC = William's Cave; C- 
08 and C-09 = High Sloth Caves; (F) = cave fill; (M) = Neotoma midden. 









WC2 


C-08 


C-08 


C-09 


Species 


Common names 


W(F) 


(M) 


(F) 


(M) 


(F) 


A. Modern local desert and scrubland species 












Agave sp. 


Century plant 






X 




X 


Artemisia cf. ludoviciana 


Estafiata 




X 






X 


A triplex canescens 


Four-wing saltbush 


X 


X 


X 






Brickellia sp. 


Brickell-bush 






X 


X 


X 


Ceanothus sp. 


Buck-brush 










X 


Cercocarpus montanus 


Mountain mahogany 






X 




X 


Chenopodium sp. 


Goosefoot 






X 






Chrysothamnus sp. 


Rabbit brush 








X 




Cucurbit a sp. 


Gourd 


X 










Dasylirion leiophyllum 


Smooth-leaf sotol 






X 






Echinocereus sp. 


Hedgehog cactus 






X 




X 


Ephedra sp. 


Mormon tea 








X 




Fallugia paradoxa 


Apache plume 






X 






Fendlera sp. 


Fendler-bush 










X 


Garry a ovata 


Silk-tassel 






X 






Helianthus sp. 


Sunflower 


X 


X 


V 






Lappula sp. 


Stick-seed 


X 








X 


Lesquerella sp. 


Bladder-pod 








X 




Lithospermum sp. 












X 


Merit zelia sp. 


Stick-leaf 




X 








Mortonia scabrella 


Tick-weed 






X 




X 


No Una sp. 


Beargrass 






X 






Oenothera sp. 


Evening primrose 






X 






Opuntia imbricata 


Cane cholla 




X 


X 




X 


O. (Platyopuntia) sp. 


Prickly pear cactus 




X 


X 




X 


Panicum cf. arizonicum 


Panic grass 






X 




X 


Phacelia sp. 


Wild heliotrope 




X 






X 


Phoradendron sp. 


Mistletoe 




X 








Physalis sp. 


Ground-cherry 




X 






X 


Quercus pungens or 














undulata 


Scrub oak 






X 




X 


Quercus sp. 


Oak 


X 


X 








Rhus sp. 


Sumac 


X 


X 






X 


Sphaeralcea sp. 


Globemallow 




X 


X 






Yucca sp. 


Yucca 


X 




X 




X 


B. Relict species in the modern flora near sites 












Berberis haematocarpa 


Algerita 






X 


X 


X 


B. trifoliolata 


Barberry 


X 


X 








Pinus edulis 


Colorado pinyon 


X 


X 


X 


X 


X 


C. Woodland species present today at higher elevat 


ions in 


Guadalupe 


: Mountains 




Celtis reticulata 


Net-leaf hackberry 




X 


X 






Juniperus sp. 


Juniper 


X 


X 


X 


X 


X 


Ostrya knowltonii 


Hop-hornbeam 




X 


X 


X 


X 


Pinus strobiformis 


Southwestern white 














pine 






X 


X 


X 



PLEISTOCENE PLANTS 17 

TABLE I. (continued) 







WC2 


C ON 


C ok 


C 09 


Species 


COmmon names W(F) 


(M) 


(F) 


(M) 


(F) 


Prunus serotina 


Black cherry 


X 








Pseudotsuga menziesii 


Douglas fir 




X 


X 


X 


Quercus gambelii 


Gambel oak 




X 




X 


Robinia neomexicana 


New Mexican locust 


X 


X 


X 


X 


D. Extralocal species not 


present in the modern flora of 










Guadalupe Mountains 










Arctostaphylos sp. 


Manzanita 




X 




X 


Juniperus communis L. 


Dwarf juniper 






X 




Pice a sp. 


Spruce 






X 


X 


Rubus strigosus Michx. 


Raspberry 








X 



habitats of New Mexico, Arizona, the Rocky Mountain states, and extends 
southward into the Sierra Madre Occidental of Chihuahua and Sonora, 
Mexico. Its absence in the Guadalupe Mountains today is difficult to 
explain, but may indicate a hot, dry stress period during the Holocene. The 
nearest population of Juniperus communis is 325 km to the northwest in the 
southern end of the Rocky Mountains in north-central New Mexico (Fig. 3). 

Arctostraphylos sp. could represent either A. pungens (point-leaf man- 
zanita) or A. uva-ursi (bear-berry manzanita). A. pungens is a widespread 
southwestern species characteristic of chaparral communities. The only rec- 
ord of it in Trans-Pecos Texas is in the Davis Mountains, 1 10 km to the 
southeast (T. L. Burgess, pers. comm. 1975). A. wvtf-wrs/isalowshrubinthe 
understory of subalpine forests from the Rocky Mountains south into north- 
central New Mexico (Vines 1960). The remainder of the forest species in the 
W 1 A Zone presently occurs in the Guadalupe Mountains, but is restricted to 
mesic habitats such as McKittrick Canyon and The Bowl (upper Pine 
Springs Canyon) on the other side of the escarpment (Gehlbach 1967). 

Wisconsin IB. — Williams Cave is located at 1500 m elevation on the 
south-facing bajada below El Capitan Peak on the south end of the 
Guadalupe Mountains. An ancient packrat midden in Williams Cave con- 
tained a rich pinyon-juniper macrofossil assemblage including Pinus edulis, 
Juniperus sp., Robinia neomexicana, Prunus serotina (black cherry), Celtis 
reticulata (netleaf hackberry), and Quercus sp. (Table 1). A radiocarbon 
date on Juniperus sp. twigs and seeds was 12,0 10 ± 210 BP (A- 1540). A pol- 
len assemblage from a roughly contemporaneous cave fill sample from Wil- 
liams Cave contained very high Juniperus, moderate Pinus, high total ar- 
boreal pollen (DAP; 82.5 to 90.5%), and moderate Artemisia pollen (Fig. 2, 
Table 2). 

The fossil Juniperus material probably represents J. monosperma (one- 
seed juniper), but J. pinchotii (red-berry juniper) cannot be ruled out on the 
basis of leaf morphology. Both J. deppeana (alligator-bark juniper) and J. 



18 



VAN DEVENDER ET AL. 







1 



E 



2 

O 



J 



o 

C 

g-Ji 



o 8 



3 fl 



. 00 



oj ro *j- in ^) 



PLEISTOCENE PLANTS 



Pseudostugo m e n / 1 e s i 




Rubus strigosus 




P i c e a e 


n 


9 


e 1 m a 


n n i i 










• 
*• 
• - 


1 




) % 

• 










• 






*( 


• 




• 
• 


f 




• 




1 








^"\^^ • 













Fig. 3. Distribution maps of important species in the late Pleistocene plant macro- 
fossil assemblages in the Upper Sloth Caves, Guadalupe Mountains, Culberson 
County, Texas. Maps A, C and D after Little (1971). Map B from University of 
Arizona Herbarium specimens; probable areas are stippled. 



scopulorum ( Rocky Mountain juniper) can be identified, but were not in the 
Williams Cave deposits. 

The present vegetation near Williams Cave is Chihuahuan desertscrub 
with shrub and succulent components. Important shrubs include Larrea di- 
varicata (creosote-bush), Acacia neovernicosa (viscid acacia), Mortonia 
scabrella (sandpaper bush, tick-weed), Fouquieria splendens (ocotillo), 
Prosopis glandulosa (Torrey mesquite), and Viguiera stenoloba (skeleton- 
leaf goldeneye). Important succulents include Agave lecheguilla (leche- 
guilla), Dasylirion leiophyllum (sotol), Yucca baccata (banana yucca), Y. 
?lata (soapweed yucca), Y. torreyi (Torrey yucca), and many cacti in the 



20 



VAN DEVENDER ET AL. 



I | 

CD X> 

•° £ 

cd C 



ft 


«l 


P9 







£> 




cd 


73 


-J 


^ , 




CD 


II 



f- < 



- J a 

II 

o ^ 

31 



§ ii 



Us 

"e3 o 

cd . 

2 * 

O £ 

a, o 

4S ,o 



5 II 

.2 ^ 

"cd j> 

S? r i 



•a ^ 

O II 

o ^ 
IS i? 



£ ii 
"o c/a 



CD II 



s p < 



« II 



hJ K 42 



3 2 J= 

PQ ft D 



o ^ 

J=> ft 

o 

•3 



o & 
'•3 o 





^— 


03 


CD 


^6 


O 


cd 
£ 


ft 

>, 


co 


^ 


£ 


CO 

3 


o 


.ap 


cd 


1 

U 


ft 
< 


o 

cd 


"S 1 












(D 


"rd 










VO 




ed 


03 




C^ 


*H 


T3 




O 

I 

U 


CD 

o 
£ 


O 

£ 
o 


co 

o 

1 

U 












a 

-o 


^ 


j=. 




o 


Mi 


_<Uj 


Ch 








£ 


s; 


ic 


2 



C£ o . 

i If 

W ^ rP £ 

CD VO * O 

2 •• c » 

cd °° cd i3 

^ 9 «« -a 

C w cd ft 







«3 


.^_* 


















69 

3 


3 


o 


"53 






II 






69 

3 


£ 


5a 

•2 


, 6 ' 


69 


2 


3 




Cj 


2 


3 


5 


p 


u 


^3 




•2 


ft 


63 


2 




^J 


£ 




2 


C2 


3 




69 

o 


S- 


5 


,=o 


_ 





?s 


s 


b<3 

09 




g 


5j 

•2 


o. 
g 

^ 


ft 

co 


"3 
3 




1 




-5 




g 


§ 


69 


-5 


os; 


^ 



o 




o 




CO 




m 




CN 




t> 




+1 


^^ 


+i 




O 


CM 

CD 


o 


&9 

CD 


MD 




o 




O 


-o 


o 


S 




CD 




CD 


CO 


CD 


CO* 


>D 


rH 


C 


— i 


c 




-3 




«3 


-f 




CO 




cr> 


g 


l^ 


5, 


<-H 


1 ' 


i-H 




< 




< 




CO 


CTn 






O 


o 






U 


U 







.S.2-& 

N 6, ■ " 





^j 


53 
CD 
5* 


C3 


o 


CD 


69 


CD 

jc: 
U 


CD 

£ 


CO 


o 


^ 


cd 




73 


_o 


6 


do 
o 


O 


■ 2 




1 


CO 


69 


^ - 


U 


rt 


5 


CO 


co 

CD 


■a 



ft CD 



6) ^ 







-2 

3 

■2 


1 






ft 

CO 

CO 


s 


C3 






3 


,69 


.CD 
















CD 




5-5 






■S- 


2 


CD 






3 


r 









s 


ft 


<M 










K 




Co 

3 

CO 


^ 


CO 

3 


.2 


2 


1 


CD 


s 





v| 









V, 


Co 


3 


0)C< 


69 




K 










»: 









ri 








+1 


CO 


O 


'it 


O 


* j 




69 


rsT 

1 — 1 


3 




■& 









~ 


-t 


a 


U-) 


^ 


— ' 




< 





u .= 



on 



- £ 

2 2 I 

CD ^ " 

o o g £ 

p. £ ^O 

< CJj -i 

X) 6 n cd* 

C S -^ 2 

rt « CO ed 

co r* 'C 

ij* x 2 ^ 2 

Tl ^0 CD j3 CD 

x !3 *o --2 "O 

wj « ° S ° 

2 2 £ a £ 







69 
3 

Cj 








A 


ik 


C3 








^ 


,«j 




69 
3 


1 


3 


cu 

s 




K 




. ~ 




s; 


/■-'' 


K 




CD 






"3 


O 


K 




69 

-2 


CD 


"S 


.2 




N 


^ 





.s; 


a 
*« 




K 


-5 


^ 


5j 


,69" 


-<: 





3 

.CD 


s 




a 


c< 


1 


69 


~2 


d 

s 


CO 




-Q 


O 


~q 


■2 


? 


^ 


8. 


d: 


3 


69 


CO 


a 



o 

CS 

+1 



o 



o o 

+1 3 +1 +1 +1 
o o " T3 o o o 



2 b^jt 

IT) 00 IO 



CO 

<co 



w JS CO — CO -5 
lO OO 10 c>0 10 00 



PLEISTOCENE PLANTS 



21 



2 °< 

43 O 



o 
IS £ c 

„ O j) 

-as 



cc 



O "3 « 






Cfl 



"8 t 

5o * C * 

0) ^j <U on 



03 03 +-< 



r3 



5T o 

S-c s" 
6, a a 
a 



•I a 

5 5» 








o 2 



O Q 



U U 



go "> 

o o 

o o 

i/-) o 



22 VAN DEVENDER ET AL. 

Coryphantha, Echinocereus, Mamillaria, and Opuntia genera. A few relict 
Juniperus monosperma are in a sandy arroyo 1 km to the east at the same 
elevation as Williams Cave. 

Presently, pinyon-juniper woodland probably analogous to the paleo- 
community recorded in the Williams Cave packrat midden occurs as low as 
1550 to 1700 m in the Pine Springs Canyon and Frijole Ranch areas. How- 
ever, Quercus is more important in these areas than it was in the fossil com- 
munity. Yet, the south face of El Capitan Peak above Williams Cave is so 
steep and rocky that a modern analogue to the paleocommunity might be 
over 2450 m elevation. The physiography of the Guadalupe Escarpment is so 
complex that we will not attempt to estimate the depression of vegetation 
zones due to Pleistocene climates. 

Although the Williams Cave samples are intermediate in age between the 
two periods recorded in the Upper Sloth Caves, we feel that the ancient 
pinyon-juniper community probably corresponds to the more mesic of the 
two, Zone Wl. 

Wisconsin 2. — Trench 1 in Cave C-08 has a multilayered stratigraphy with 
both Holocene and late Pleistocene units. Stratum 3 (25 to 45 cm) contained 
an upper layer of dung of the extinct Shasta ground sloth and of wood 
(Lloyd E. Logan, pers. comm. 1975). Directly beneath this unit was a mat of 
leaf litter designated as Stratum 3a (40 to 45 cm). A large sample of the leaf 
litter was collected, sorted, and identified (Table 1). The macrofossils from 
Stratum 3a document a mixed conifer forest composed of Pseudotsuga men- 
ziesii, Pinus strobiformis, P. edulis, Juniperus sp., Ostrya knowltonii, 
Quercus gambelii, Robinia neomexicana, and Celtis reticulata. A similar 
stratigraphic section was seen in a nearby pit from a previous excavation. A 
radiocarbon date on large artiodactyl fecal pellets (possibly Cervus elaphus 
merriami, Merriam's elk) from the equivalent of Stratum 3a, Trench 1 in that 
pit (below the sloth dung layer) is 1 1,760 ±610 BP (A- 1533). A date on sloth 
dung from Cave C-05 is 1 1 ,590 ± 230 BP ( A- 1 5 1 9), which demonstrates that 
the sloths inhabited the caves during Zone W2. The pollen samples from this 
zone are characterized by absent to moderate Picea, moderate to high Pinus 
and S AP, high Ostrya, low to moderate Cheno-ams (Chenopodiaceae plus 
Amaranthus), and moderate Gramineae (Fig. 2, Table 2). 

Zone W2 differs from Zone 1 A (Caves C-08 and C-09) in the absence of 
the extralocal mesic forest species from the immediate vicinity of Cave C-08, 
i.e., Picea sp., Juniperus communis, Rubus strigosus, and Arctostaphylos 
sp. The difference in climate needed to account for this absence may not 
have been too dramatic. Spruce pollen is not transported as far, nor is 
spruce as prolific a pollen producer as pine (Potter and Rowley 1960). The 
presence of small, but significant amounts of Picea pollen suggest that 
spruce may have been restricted to mesic areas either in front of Caves C-05 
and C-09 (north-facing) or in mesic habitats on the east side of the mountains 
rather than being completely extirpated from the area. Also, a skull of Sorex 
cinereus (masked shrew), a subalpine forest species presently living no closer 



PLEISTOCENE PLANTS 



23 



TABLE 3. Holocene and late Pleistocene plant macrofossils associated with pollen samples 
from Caves C-08 and C-09, Guadalupe National Park, Culberson County, Texas. Pollen 
data, depth of samples, and stratigraphy are presented in Fig. 2. 



90 90 

O © 

I I 

u u 



Species 








A. Modern local desert and scrubland 


species 






Agave sp. 






X 


Atriplex canescens 




X 


X 


Cercocarpus montanus 


X 




X 


Dasylirion leiophyllum 








Echinocereus sp. 


X 






Fallugia pa f adoxa 


X 


X 




Garrya ovata 






X 


Kallstroemia parviflora 




X 




Mortonia scabrella 


X 






Oenothera sp. 








Opuntia (Platyopuntia) sp. 


X 


X 




Panicum cf. arizonicum 








Quercus pungens or undulata 






X 


Sphaeralcea sp. 








Yucca sp. 


X 




X 


B. Relict species in modern flora near sites 






Berberis haematocarpa 






X 


Pinus edulis 


X 




X 



C. Woodland species present today at higher elevations 
in Guadalupe Mountains 

Celtis reticulata X 

Juniperus sp. 

Ostrya know It on ii 

Pinus strobiformis X 

Pseudotsuga menziesii X 

Quercus gambelii X 

Robinia neomexicana 

D. Extralocal species not present in the modern flora 
of Guadalupe Mountains 

Arctostaphylos sp. X 

Juniperus communis 
Picea sp. 
Rubus strigosus 



X X 

X 



X X 

X 
X 



than northern New Mexico, was found in Stratum 3a with the Zone W2 
assemblage ( Lloyd E. Logan, pers. comm. 1975). Minor changes of this sort 
in the vegetational composition would be expected with a gradually warm- 
ing climate. The most dramatic change in the vegetation was probably from 



24 



VAN DEVENDER ET AL. 



TABLE 4. Pollen counts of samples from Cave 09, Test Trench II (C-09), Cave 08, Trench 1 (C- 
08), Cave 05, Trench 6 (C-05) and Williams Cave (WC). Depth of samples and stratigraphy 
presented in Fig. 2. 





- 


c^ 


_ 


m 


Ci 


•^ 


id 


v© 








Pollen types 


© 


o 


90 

© 


90 

9 


90 

© 


90 

O 


90 

© 


90 

© 


© 


(J 


u 




U 


U 


U 


u 


U 


U 


u 


U 


u 


£ 


£ 


Picea 


8 


7 








4 




3 




1 




Pseudotsuga 


1 














1 








Pinus 


92 


34 


22 


34 


13 


52 


22 


26 


21 


14 


51 


Juniperus 


10 


13 


12 


3 


17 


17 


7 


12 


4 


150 


130 


Ostrya 


26 


50 






19 


30 


3 


6 


8 




1 


Quercus 


7 


25 


14 


14 


13 


5 


16 


12 


13 






Alnus 






1 


















Juglans 




1 






1 






1 








Celtis 












1 












Sarcobatus 
















1 








Cheno-ams 


1 


2 


46 


42 


34 


5 


37 


12 


38 


1 


3 


Gramineae 


4 


4 


7 


9 


19 


10 


19 


32 


29 






Ephedra torreyana-iypt 




1 


5 


24 


1 




5 


3 








E. nevadensis-type 














1 








1 


Artemisia 


4 


4 


13 


14 


11 


19 


13 


22 


8 


12 


3 


Short-spine Compositae 




2 


17 


13 


11 


4 


8 


2 


3 


1 




Long-spine Compositae 


24 


12 


28 


20 


31 


31 


28 


39 


35 


12 


3 


Senecio-Xypc 


















1 






Cercocarpus-type 


9 


16 


16 


9 


6 


3 


14 


10 


8 


2 


2 


Rubus-iype 


1 


3 




















Eriogonum 






3 


2 


3 














Ericaceae 






1 


















Cruciferae 




1 












2 


1 


1 




Scrophulariaceae 


1 


1 








2 












Onagraceae 


















1 






Crossosoma-type 
























Leguminosae 


1 


1 


1 


1 


2 




4 




3 


1 




Ribes 
















1 








Saxifragaceae 


1 


6 




1 




1 












Rhus 




1 












2 






1 


Cereus-type 


1 






















Tide sir omia 














1 










Caryophyllaceae 








1 




1 




2 


2 






Nyctaginaceae 














1 




1 


1 




ISambucus 














1 










Euphorbia 








1 






1 










Malvaceae 












1 


2 










Yucca 












1 












Primulaceae 








1 
















Undetermined 




5 


3 


2 


4 




2 


2 


7 


2 


1 


Unknown 


9 


11 


11 


9 


10 


13 


13 


10 


17 


3 


4 


N 


200 


200 


200 


200 


200 


200 


200 


200 


200 


200 


200 


Sap 


144 


130 


49 


51 


63 


108 


48 


61 


46 


165 


182 


Trilete, psilate spores 


2 


6 


19 


20 


9 


3 


12 


14 


2 






Lycopodium-lype spores 










1 








17 







PLEISTOCENE PLANTS 25 

Zones W2 to HI (Holocene 1) accompanying the climatic change that 
marked the beginning of the Holocene. 

Holocene 1. — Several undated stratigraphic units in Caves C-05 and C-08 
appear to be early Holocene in age. The plant macrofossils are mostly xero- 
phytic species that presently grow near the caves. Juniperus sp., Celtis re- 
ticulata, and perhaps Pinus edulis (some possibility of contamination be- 
cause the sample contained but a single needle) are the only remnants of the 
former forests that grew near the caves. The climatic sequence during the 
Holocene is controversial for that portion of the Southwest where summer 
monsoonal rain is important (western Texas, southeastern Arizona, and 
New Mexico); r(Antevs 1962; Freeman 1972; Johnston 1963). The Antevs 
model of a warm, moist Anathermal period from 9000 to 7000 BP followed 
by a hot, dry Altithermal period was based on geological and palyno- 
logical data from the Great Basin and a chronology extrapolated from the 
Swedish literature (Antevs 1952). That climatic sequence has been general- 
ly, and often uncritically, applied to the summer monsoon areas of the 
Southwest. However, Martin (1963a) and Mehringer et al. (1967) inter- 
preted the pollen record for southeastern Arizona as indicating a warm, dry 
climate from 10,000 to 8000 BP and a warm, moist climate from 8000 to 
3000 BP. Recently, Van Devender and Worthington (1977), using fossil 
reptile and amphibian data from Howell's Ridge Cave, Little Hatchet 
Mountain's, southwestern New Mexico, suggested that both the Ana- 
thermal and the Altithermal periods were warm and moist and that the 
nearby playa did not dry up until about 4000 years ago. Perhaps Juniperus 
sp. and Celtis reticulata persisted near the Upper Sloth Caves in the Guada- 
lupe Mountains during a warm but relatively moist period in the early 
Holocene. Such a climatic regime might explain the unusual occurrence of 
mesic reptiles such as Phrynosoma douglassi (short-horned horned lizard) 
and Thamnophis sp. (gartersnake) in Holocene deposits in Pratt Cave, 
McKittrick Canyon, on the east side of the Guadalupes (Gehlbach and 
Holman 1974). 

Holocene 2. — This zone contains modern vegetation and suggests a mod- 
ern climatic regime. It differs from H 1 only in the absence of the xero- 
philous woodland species such as Juniperus sp., Celtis reticulata, and possi- 
bly Pinus edulis. The discussion above on the Altithermal controversy shows 
that the Holocene vegetation and climate are more complex than these two 
units would imply. Undoubtedly, a great many details remain to be learned. 

DISCUSSION 

The late Wisconsin plant communities have been discussed in a surprising 
number of publications, considering the scarcity of plant fossils. Stearns 
(1942) and Murray (1957) suggested that the late Pleistocene life zones were 
lowered 1230 m and 1075 m, respectively, based on the present distributions 
of such subalpine mammals as Marmota flaviventris and Neotoma cinerea. 
Both authors believed that the Hudsonian or Canadian life zones (= 



26 VAN DEVENDER ET AL. 

spruce-fir forest) formerly grew at 1400 m elevation in the vicinity of Burnet 
Cave, 50 km W of Carlsbad, Eddy County, New Mexico (Schultz and 
Howard 1935). Antevs (1955) suggested a 770-m depression of life zones at 
Burnet Cave, using estimates of snowline depression derived from cirque 
elevations of late Pleistocene montane glaciers in Colorado and New 
Mexico. The glacier nearest to the Guadalupe Mountains was on Cerro 
Blanco (= Sierra Blanca), 70 km W of Roswell, New Mexico. Antevs 
assumed that the subalpine mammals were transported to Burnet Cave from 
forest habitats on the nearby mountain crests which are at 2000 m elevation 
in this part of the Guadalupes. Recently, Galloway (1970) suggested a 1300 
to 1400 m late Pleistocene lowering of timberlines in the entire Southwest, 
using periglacial solifluction deposits in the Sacramento Mountains just 
north of the Guadalupes. He assumed that the upper treeline was no higher 
than 2050 m and arrived at a paleoclimate of 10 to 1 1° C lower mean annual 
temperature than the present and 80 to 90% of the present precipitation. This 
scheme would place the lower edge of the Douglas fir-southwestern white 
pine forest, presently above 2400 m in the Guadalupe Mountains (Gehlbach 
1967), along the Pecos River to the east at ca. 1075 m elevation. However, 
both Antevs (1955) and Leopold (1951) concluded that pluvial Lake 
Estancia, New Mexico, needed increased precipitation as well as reduced 
temperature to maintain it. 

Taking Zone Wl A in the biochronological sequence presented above, let 
us examine these ideas. Picea sp., Pseudotsuga menziesii, Pinus 
strobiformis, and other montane species existed at an elevation of 2000 m on 
the exposed, steep, relatively xeric west side of the escarpment 13,000 years 
ago. The present vegetation zones on the east side of the escarpment are 
lower, and probably the Zone Wl A subalpine forest extended as low as 1700 
m on that side where there is a gentler slope and a more mesic habitat. This 
forest probably occurred to at least 2475-m elevation above the Upper Sloth 
Caves. In the light of this evidence, as well as that of others (Wright et al. 
1973), we reject the values for timberline depression given by Galloway 
(1970). 

The Williams Cave record (Zone W 1 B) demonstrates that on the bajada at 
the southern end of the Guadalupes at 1 500 m elevation the paleocommu- 
nity was a pinyon-juniper woodland (the Juniperus- Pinus dominance type 
of Gehlbach 1967). The pollen record from this zone has a very high 2 AP 
dominated by Juniperus. We interpret this as representing a woodland 
dominated by Juniperus (probably /. monosperma) with scattered Pinus 
edulis. The fossil pollen assemblages contained no record of the present 
Chihuahuan desertscrub vegetation. A xerophilous juniper woodland 
probably extended west to the Salt Flat playa at 1 100 m and east to the Pecos 
River at 1075 m elevations. This reconstruction is supported by Wells' (1966) 
record of xerophilous woodland as low as 615 m in the Big Bend of Texas, 
330 km to the southeast. We feel that a pinyon-juniper woodland extended 
northward along the eastern flanks of the Guadalupe Escarpment at similar 



PLEISTOCENE PLANTS 27 

elevations at least to the Burnet and Dry Cave areas ( 1 290 to 1 420 m) west of 
Carlsbad. Today, the woodland extends lower on the east side of the 
Guadalupes, and the late Pleistocene community of this area was probably 
somewhat more mesic than that of the Williams Cave area. The distribution 
of Pinus ponderosa (ponderosa, or yellow pine) at that time is not known, as 
we have not found macrofossils of this species in our sites. It may not have 
expanded considerably beyond its present range. At any rate, the vegeta- 
tion near Burnet Cave was not the Hudsonian or Canadian Zone forests sug- 
gested by Stearns (1942) and Murray (1957). 

Based on the faunai record, Harris (1970) reconstructed the paleo- 
environment of the Dry Cave area as Transition Zone big sagebrush 
{Artemisia tridentata) and grassland communities with scattered junipers 
and yellow pines. Lagurus cur tat us (sagebrush vole) is a member of the Dry 
Cave fauna that is presently restricted to the big sagebrush communities 
farther north. Artemisia tridentata presently occurs no closer than north- 
central New Mexico. The Williams Cave and Upper Sloth Caves pollen 
samples did contain low to moderate amounts of Artemisia pollen, but A. 
tridentata is a prolific pollen producer (see discussion in Martin 1963b). A. 
ludoviciana (estafiata) is a widespread, common, herbaceous species that 
presently grows near both sites and was also a component of the Pleistocene 
macrofossil assemblages from the Upper Sloth Caves. The Williams Cave 
samples do not record a dense A. tridentata community and may only 
represent A. ludoviciana. If A. tridentata were present near Williams Cave, it 
must have been as widely scattered individual plants. The Williams Cave 
samples contained no grass pollen at all. Hence, the big sagebrush or grass 
reconstruction cannot be sustained for the Williams Cave area. Dry Cave is 
about 55 km NNE at 1290 m elevation and it is possible that A. tridentata 
was better represented there. At any rate, the paleocommunity in the Dry 
Cave area was probably an Upper Sonoran Zone pinyon-juniper wood- 
land rather than a Transition Zone forest, grassland, or sagebrush com- 
munity. 

The Zone W 1 A forest on the top of the Guadalupe Mountains was a mix- 
ture of Hudsonian (Picea sp., Juniperus communis), Canadian (Pseudo- 
stuga menziesii, Pinus strobiformis), Transition (Quercus gambelii, Ostrya 
knowltonii), and Upper Sonoran (Pinus edulis, Juniperus sp.) Life Zone 
elements. Abies concolor (white fir) and A. lasiocarpa (cork-bark fir) are 
common subalpine forest, mixed-conifer, or pine forest species that were not 
found in any of the fossil samples, nor was Abies pollen found in any of our 
Guadalupe Mountains Pleistocene samples. Abies pollen cannot be trans- 
ported great distances, but if Abies were at all common in the Pleistocene 
communities, some pollen grains would be expected. Stands of Pinus pon- 
derosa, an important southwestern Transition Zone species, presently occur 
as low as 1 700 m on the east side of the escarpment. It was not found in any of 
the Pleistocene macrofossil samples, although it was surely in the moun- 
tains at that time. The Zone Wl A forest was a diverse mixed-conifer forest 



28 VAN DEVENDER ET AL. 

that cannot be fitted easily into the Merriam Life Zone system. The 
responses of these species were individualistic in nature as was suggested by 
Gleason(1939). 

If plant communities are considered as coincident, overlapping distribu- 
tions of plant species along several environmental gradients (including 
a time gradient; Mcintosh 1958, 1967; Whittaker 1967), the relationships of 
plant communities to climate must be general, i.e., the adaptations to climate 
are at the species level. Most animals have behavioral adaptations which 
help them to ameliorate extremes of temperature or moisture. Animals 
generally have a wider ecological amplitude than plants, and even the sub- 
alpine species in the Pleistocene faunas (Marmotaflaviventris and Neotoma 
cinerea) are found occasionally in lower-elevation habitats (Harris 1970; 
Finley 1958). Perhaps the fauna and flora together should be viewed as a 
biotic continuum in which animal species occasionally lived in somewhat 
different habitats than at present. Marmota flaviventris and Neotoma 
cinerea may well have been more common in late Pleistocene 
pinyon-juniper woodlands than they are at present. An important con- 
sideration in the reconstruction of past environments using present biotic 
distributions is that the present climate is probably unusual. Most of the 
biota have endured a greater period of time under the Pleistocene glacio- 
pluvial climates than under the present interglacial conditions. 

LITERATURE CITED 

Antevs, E. 1952. Cenozoic climates of the Great Basin. Geol. Rundschau 40:94- 108. 

1955. Climate of New Mexico during the last glacio-pluvial. J. Geol. 

62:182-191. 
1962. Late Quaternary climates in Arizona. Am. Antiq. 28:193-198. 



Ayer, M. Y. 1936. The archaeological and faunal material from Williams Cave, 
Guadalupe Mountains, Texas. Proc. Acad. Nat. Sci. Phila. 88:599-618. 

Correll, D. S.,and M. S. Johnston. 1970. Manual of the Vascular Plants of Tex- 
as. Texas Res. Found., Renner, Texas, 1881 pp. 

Finley, R. B.,Jr. 1958. The wood rats of Colorado: distribution and ecology. Univ. 
Kans. Mus. Nat. Hist. Publ. 10:213-552. 

Freeman, C. E. 1972. Pollen study of some Holocene alluvial deposits in Dona Ana 
County, southern New Mexico. Tex. J. Sci. 14: 203-220. 

Galloway, R. W. 1970. The full-glacial climate in the southwestern United States. 
Ann. Assoc. Am. Geog. 60:245-256. 

Gehlbach, F. R. 1967. Vegetation of the Guadalupe escarpment. New Mexico- 
Texas. Ecology 48:404-419. 

Gehlbach, F. R., and J. A. Holman. 1974. Paleoecology of amphibians and rep- 
tiles from Pratt Cave, Guadalupe Mountains National Park, Texas. Southwest. 
Nat. 19:191-197. 

Gleason, H. A. 1939. The individualistic concept of the plant association. An\. 
Midi. Nat. 21:92-110. 

Gordon, A. G. 1968. Ecology of Picea chihuahuana Martinez. Ecology 49:880-896. 

Harris, A. H. 1970. The Dry Cave mammalian fauna and late pluvial conditions in 
southeastern New Mexico. Tex. J. Sci. 22:3-27. 



PLEISTOCENE PLANTS 29 

Howard, E. B. 1932. Caves along the slopes of the Guadalupe Mountains. Tex. 
Arch. Paleontol. Soc. Bull. 4:7-19. 

Johnston, L., J R. 1963. Pollen analysis of two archaeological sites at Amistad Res- 
ervoir, Texas. Tex. J. Sri. 15:225-230. 

Little, E. L. Jr. 1971 Atlas of United States trees. Vol. 1: Conifers and Important 
Hardwoods. U.S. Dep. Agric. Misc. Publ. 1146:1-12, 94 maps. 

LEOPOLD, L. B. 1951. Pleistocene climate in New Mexico, Am. J. Sci. 249:152-168. 

Logan, L. E., and C. C. Black. 1977. The Quaternary vertebrate fauna from the 
west side guano caves, Guadalupe Mountains National Park, Texas. This volume. 

McIntosh, R. P. 1958. Plant communities. Science 128:115-120. 

1967. The continuum concept of vegetation. Bot. Rev. 33:130-187. 

Martin, P. S. 1963a. The Last 10,000 Years: A Fossil Pollen Record of the Ameri- 
can Southwest. Univ. Arizona Press, Tucson, 87 pp. 

1963b. Geochronology of pluvial Lake Cochise, southern Arizona. II. Pol- 
len analysis of a 42-meter core. Ecology 44:436-445. 

Mehringer, P. J., Jr., P. S. Martin, and C. V. HaynesJ r. 1967. Murray 
Springs, a mid-postglacial pollen record from southern Arizona. Am. J. Sci. 
265:786-797. 

M urray, K. F. 1957. Pleistocene climate and the fauna of Burnet Cave, New Mex- 
ico. Ecology 38:129^132. 

Potter, L. D., and J. Rowley. 1960. Pollen rain and vegetation, San Augustin 
Plains, New Mexico, Bot. Gaz. 122:1-25. 

Schultz, C. B., and E. B. Howard. 1935. The fauna of Burnet Cave, Guadalupe 
Mountains, New Mexico. Proc. Acad. Nat. Sci. Phila. 87:273-298. 

Stearns, C. E. 1942. A fossil marmot from New Mexico and its climatic signifi- 
cance. Am. J. Sci. 240:867-878. 

Van Devender, T. R., and R. D. Worthington. 1977. The herpetofauna of 
Howell's Ridge Cave and the paleoecology of the northwestern Chihuahuan 
Desert. In R. H. Wauer and D. H. Riskind, eds., Transactions-Symposium on 
the Biological Resources of the Chihuahuan Desert Region, U.S. and Mexico. 
National Park Service, Washington, D.C., in press. 

Vines, R. A. 1960. Trees, Shrubs and Woody Vines of the Southwest. Univ. Texas 
Press, Austin, 1104 pp. 

Wells, P. V. 1966. Late Pleistocene vegetation and degree of pluvial climatic 
change in the Chihuahuan Desert. Science 153:970-975. 

Wright, H. E., Jr., A. M. Bent, B. S. Hansen, and L. J. Maher, Jr. 1973. Present 
and past vegetation of the Chuska Mountains, northwestern New Mexico. Geol. 
Soc. Am. Bull. 84:1155-1180. 



ACKNOWLEDGMENTS 

Paul S. Martin in his continuing pursuit of the Shasta ground sloth 
{Nothrotheriops shastense) led us into the Guadalupes and initiated the 
present study. Lloyd E. Logan and Tony L. Burgess, Texas Tech University, 
Jim I. Mead, University of Arizona, and Ben and Cyndi Everitt, El Paso 
Archeological Society, helped in the field work. Gary Ahlstrand, Roger 



30 VAN DEVENDER ET AL. 

Reisch, Phil Van Cleave, and John Chapman of the National Park Service 
were of invaluable help in many ways. Tony L. Burgess and David K. 
Northington provided valuable information on floral distributions in the 
Guadalupe Mountains. Betty M. Fink aided in editing and typing the manu- 
script. Austin Long and Paul Damon, Radiocarbon Laboratory, University 
of Arizona, provided radiocarbon dates with support from NSF grants DEB 
75-13944 to Paul S. Martin and DEB 76-19784 to Thomas R. Van 
Devender, Department of Geosciences, University of Arizona. 

Contribution No. 606, Department of Geosciences, University of 
Arizona. 



Preliminary Report on the Ecology of 
Fire Study, Guadalupe Mountains and 
Carlsbad Caverns National Parks 



GARY M. AHLSTRAND, Carlsbad Caverns and Guadalupe 
Mountains National Parks, New Mexico 

The National Parks Act of 1916 states that the purpose of the national 
parks "is to conserve the scenery and the natural and historic objects and the 
wild life [sic] therein and to provide for the enjoyment of the same in such a 
manner and by such means as will leave them unimpaired for the enjoyment 
of future generations." Although the National Park Service has been largely 
successful in operating park lands for the enjoyment of the public, preserva- 
tion attempts have oftentimes impaired these natural areas by bringing 
about unplanned and undesired changes in the ecosystems (Stone 1965). For 
example, by stressing the protection of objects rather £han processes within 
ecosystems, suppression of all fires was justified easily in the parks (Agee 
1974). Results of this action are well known in some instances, but little 
understood in others at present. 

Leopold et al. (1963) drew major attention to management policies in the 
national parks and recommended as a primary goal "that the biotic associa- 
tions within each park be maintained, or where necessary recreated, as 
nearly as possible in the condition that prevailed when the area was first 
visited by the white man." 

The Leopold Committee (Leopold et al. 1963) urged that wildfire be 
restored to its role as an ecological factor where practical and suggested the 
use of controlled burns as a tool for restoring natural ecologic conditions to 
some national park areas. Based largely upon recommendations contained 
in the Leopold Committee Report, a new fire management philosophy has 
evolved in the past decade. The present administrative policies for natural 
areas (U.S. Department of the Interior, National Park Service 1970) state: 

The presence or absence of natural fires within a given habitat is recognized as 
one of the ecological factors contributing to the perpetuation of plants and ani- 
mals native to the habitat. 

31 



32 AHLSTRAND 

Fires in vegetation resulting from natural causes are recognized as natural phe- 
nomena and may be allowed to run their course when such burning can be con- 
tained within predetermined fire management units and when such burning will 
contribute to the accomplishment of approved vegetation and/ or wildlife manage- 
ment objectives. 

Prescribed burning to achieve approved vegetation and /or wildlife manage- 
ment objectives may be employed as a substitute for natural fire. 

In keeping with this policy, fire management plans for Guadalupe Moun- 
tains and Carlsbad Caverns National Parks designate natural burn units in 
portions of each park. The plans stipulate that the effects of fire on vegeta- 
tion, including recovery rates, will be determined. Detailed investigations of 
the potential effects of wildfire on the interior mountain and canyon areas of 
Guadalupe Mountains National Park, especially in relation to its effect on 
relict taxa or plant communities, are called for before considering inclusion 
of these areas in a natural burn unit. In addition, a study to determine the 
feasibility of using controlled burns to restore or manipulate habitats that 
have been altered severely by human activities is required by the plan. 

HISTORICAL ASPECTS 

There can be little doubt that the vegetation in the area has undergone 
change, especially during the last century. Photographs taken at Carlsbad 
Caverns National Park less than 40 years ago give evidence that dramatic 
vegetational changes have occurred since livestock grazing was eliminated 
from the park. Verbal accounts by lifelong residents of the area have told of 
lush grasslands once existing in areas presently dominated by Larrea triden- 
tata and Acacia constricta. A report by Pope (1854) made no less than three 
references to the large pine forests covering the east slopes of the Guadalupe 
Mountains in the vicinity of the Pinery. Another reference in the Pope ( 1 854) 
report told of the great abundance of grama grass surrounding a camp 
located between the salt flats and the west escarpment. 

No one factor can account for all the observed and suspected vegetational 
changes. The effects of livestock grazing, past fire history, water table fluc- 
tuations, and possible climatic changes must be assessed. Attempts to 
determine the relative influence of each of these factors will be continued 
through searches for historical accounts and old photographs of the area, 
interviews with longtime residents of the area, tree-ring studies to detect past 
fire frequencies and possible climatic changes, and age-class determina- 
tions of trees in forested areas. 

Overgrazing has shifted the competitive advantage in favor of lig- 
nophytes and allowed former grasslands to be invaded by shrubs 
(Humphrey 1953, 1974). Weakening of the grass cover also lowered the inci- 
dence of fires which were effective in checking the spread of woody species 
(Hastings and Turner 1965). Dog-hair thickets of Pseudotsuga menziesii and 
Pinus ponder osa found in portions of the relict coniferous forest may be due 
in part to the planting of seeds by the hooves of livestock. 



FIRE ECOLOGY 33 

Many of the pines reported by Pope (1854) to inhabit the east side of the 
Guadalupe Mountains were logged for construction and fuel purposes by 
the early settlers in the area. However, this activity alone does not seem to be 
a satisfactory explanation for the distribution or paucity of reproduction 
noted for Pinus in the area today. 

The flow from Lower Pine Spring was reduced greatly following an earth- 
quake in 1931. Whether the apparent lowering of the water table can be 
attributed entirely to tectonic movements, or if man's activities in the region 
also may have influenced ground-water levels in the area needs investiga- 
tion. The 193 1 earthquake also might account for the absence of young coni- 
fers in a meadow area at 2450 m elevation, located south of Pine Springs 
Canyon. At the same elevation, but to the north of the canyon, seedling 
establishment has led to the dog-hair thickets referred to previously. 

The past role of fire in the area must be determined. Inner growth rings of 
sections from fire-scarred trees in the coniferous forest gave evidence that for 
one 150-year period, fires occurred on the average of every 25 years. The 
outer rings showed no evidence of fire for over 100 years. The sections were 
representative of but one location. To determine the past frequency of exten- 
sive burns, many more sections and cores must be examined. Accurate 
dating of the fires must await the completion of a master tree-ring 
chronology for the Guadalupe Mountains. Robinson (1969) reported that 
fires swept through the relict forest in Guadalupe Mountains National Park 
in about 1858 and again in 1908. There is evidence of the occurrence of 
smaller fires contained by natural barriers during the interim between exten- 
sive burns. Results from an analysis of the size-class distribution of conifers 
in The Bowl show that under the right set of environmental conditions the 
understory now could contribute to an extensive crown fire (Table 1). The 
same analysis indicated that Pseudotsuga menziesii has invaded more 



TABLE 1. Density by size class of conifers in The Bowl. Based on data from thirty-six 5 by 
25 m plots, mixed aspects. 











Diameter 


at breast height 










0-1 dm 




1 to 2 


2 to 3 


3 to 4 


4 to 5 




Species 


<0.5 dm 


>0.5 


>5 dm 








dm 


dm 


dm 


dm 


dm 






<1 m 


>1 m 






Tall 


Tall 














Pseudotsuga menziesii 


3411 


525 


14 


9 


12 


3 


6 


4 


Pinus strobiformis 


332 


166 


14 


19 


18 


18 


4 


2 


Pinus ponderosa 


303 


453 


184 


82 


9 


4 


4 


2 


Juniperus deppeana 


14 


12 


1 


6 


2 


3 


7 


- 


Totals 


4060 


1156 


224 


116 


41 


29 


21 


8 



34 



AHLSTRAND 



TABLE 2. Dominant species of the understory and reproductive stratum in relation to slope 
aspect in The Bowl. Understory trees (Quercus spp. and Juniperus deppeana) are 
not considered. 



Slope 


Slope 


Series 


aspect 






(°) 


<%) 




160 


16 


B-10 


195 


12 


B-lia 


225 


16.5 


B-4 


260 


20.5 


B-7 


290 


36.5 


B-12b 


320 


13 


B-l 


350 


27.5 


B-8 


40 


21 


B-3 


40 


28 


B-5 


80 


18 


B-6 


105 


16.5 


B-9C 


110 


15 


B-2 



Dominants 



Overstory 



Reproductive stratum 



Pinus ponderosa 

Mixed 

Pinus ponderosa 

Pinus strobiformis 

Pinus ponderosa 

Pseudotsuga menziesii 

Pinus strobiformis 
Pseudotsuga menziesii 

Pinus strobiformis 
Pseudotsuga menziesii 

Pinus strobiformis 
Pseudotsuga menziesii 

Pinus strobiformis 

Pseudotsuga menziesii 

Pinus strobiformis 

Pinus ponderosa 

Pinus ponderosa 
Pinus strobiformis 

Pinus ponderosa 
Pinus strobiformis 



Pinus ponderosa 
Mixed 

Pinus ponderosa 
Pseudotsuga menziesii 

Pinus ponderosa 
Pseudotsuga menziesii 

Pinus ponderosa 
Pseudotsuga menziesii 

Pseudotsuga menziesii 

Pseudotsuga menziesii 

Pseudotsuga menziesii 

Pinus strobiformis 
Pseudotsuga menziesii 

Pinus ponderosa 

Pseudotsuga menziesii 

Pinus strobiformis 

Pinus ponderosa 



a A mesic habitat for this aspect. Relatively deep soil gentle slope, and shaded, 
bpiots actually spanned two different habitat types. 
c Open canopy. 



TABLE 3. Fire summary for Carlsbad Caverns National Park, 1940-74. 









Fire ; 


»ource 












Lightning 








Human 




Five-year 
period 


Number 


Total 
acres 


Average 
acres/burn 


Number 


Total 
acres 


Average 
acres/ burn 


1940-44 


6 


164 


27 




6 


45 


7.5 


1945-49 


3 


63 


21 




13 


23 


1.9 


1950-54 


6 


190 


32 




9 


< 0.1 


<0.1 


1955-59 


1 


15 


15 




3 


<0.1 


<0.1 


1960-64 


3 


<0.1 


<0.1 




1 


3 


3 


1965-69 


3 


3 


1 




2 


8 


4 


1970-74 


9 


9150 


1017a 




2 


<0.1 


<0.1 ■ 



a Exclud : ng the 9100-acre Cottonwood Burn, the average becomes 6.6 acres. 



FIRE ECOLOGY 



35 



southernly exposures (Table 2). Apparently, this can be attributed to the 
shade afforded by the overstory. Should the canopy become more open, 
Pinus ponderosa and P. strobiformis can be expected to increase in tne 
reproductive stratum. 

For centuries fires were set in southwestern pine forests by Indians for 
hunting and waging war (Cooper 1960; cited by Hanks and Dick-Peddie 
1974). Pope (1854) wrote of the Indians setting fire to the prairie a few miles 
southeast of Guadalupe Peak. Seven days later he noted, "The young grass is 
springing up on the ground that was fired a few nights ago. The prairie still 
continues to burn; the light can be seen at a distance of 45 miles from 
camp. . . ." 

The recorded fire histories for both parks are summarized in Tables 3-4. 
Since 1960, lightning-caused fires have outnumbered those caused by 
humans. Improved surveillance during periods of extreme fire danger and 
public assumption that "fires are bad" have contributed to the recorded 
reversal in trend. In general, human-caused fires have resulted in fewer acres 
burned per fire than those originating from lightning strikes. The former 
usually have been suppressed before they could spread because of easier 
access than most lightning-caused fires. 



TABLE 4. Fire summary for Guadalupe Mountains National Park, 1960-74. 



Fire source 



Lightning 



Human 



Five-year 
period 



Number 



Total Average 

acres acres/burn 



Number 



Total Average 

acres acres/ burn 



1960-64 


1 


3 


3 











1965-69 


1 


<0.1 


<0.1 


1 


<0.1 


<0.1 


1970 74 


12 


725 


60 


2 


30 


15 



Since the natural burn policy was initiated at Carlsbad Caverns National 
Park in 1972, less than one-tenth acre has burned due to nonsuppression. 
The Cottonwood Fire of 1974 burned approximately 9 100 acres of park land 
and accounted for most of the total acreage burned during the 1970 to 1974 
period. Although much of the fire burned in the natural burn zone, con- 
tinuous suppression efforts were waged until the fire was declared out. 
Burned vegetation recovered rapidly in most areas following the fire (Figs. 1- 
2). 

Usual late spring and early summer conditions of high temperatures, low 
relative humidites and fuel moisture values, high lightning probability, and 
moderate to strong winds contribute to potentially high fire dangei situ- 
ations when combined with adequate fuel supplies. Most of the recorded 



36 



AHLSTRAND 




cd ^ 



43 


£ 


ca 


i- 


cd 


6 


Cu> 


09 


i< 


«£ 


cd 


o 


Cd 


_D 


s- 


<— 


QJ 


a. 




43 


a. 


C 


*-• 


cd 


o 


■Ji 


05 


6 

u 


09 


cd 


c 


a 


■3 


o 






-o 


<H- 


u 


c 


£. 


cd 


>■» 


& 




_c 


•8 


D< 


'cd 


Eg 


cd 


£ 


3 
v. 


.ti 


T3 


& 


09 

c3 


09 


c 


44 


C/5 


3 


cd 


c 


S 


0. 


o 


G 
O 


"cd 


(U 




c 


t-. 


43 


o 


1+h 


op 


cd 


43 


'53 

43 


Z 


>> 


43 


c/5 


4D 


CJ 


C 


■o 





U 


u 


> 


E 


09 


cd 

u 

cd 

09 


3 
c 



o 

09 


2 
> 

< 


cd 

u 

o 


a 

c 

SI 


09 

Cu> 
> 

cd 


>> 


o 


>3 


Ih 


1- 


3 


cd 


O 




c 




^> 


3 

o 

JO 

C/3 


T3 
U 

'J5 


-a 
c 


<L> 




cd 


£ 


cd 


a 


<L> 


4= 


-c: 


JC 


>> 








C 


cd 


C 


5 


u 




u 


c 


cd 


5 


c 


C 


u 


o. 


a 


P 


cd 


.K 


pq 


u 




o 




1 


o 


2 


-a 

c 


c 


« 


cd 


o 


&C 


5 

-— 


ti 


T 


o 

u 


T3 

c 


u 


cd 


"S: 


4= 


<3 


c< 


«-' 


JJJ 


,3 


o 


2 


-5 


£ 


A 


c 



4> O o 

1/5 ^N 

U. go o E 



FIRE ECOLOGY 



37 



c 


C 


<u 


<U 


-o 






cS 


"> 


D, 


u 


a, 


tt 


cd 


X 


JL> 






2 


.ti 


o 


"- < 


1_ 


xi 


bfl 


*j 


<u 


£ 








o 

i— 
o 


60 


o 


u 


c/o 


> 


g 


£ 


O 


TJ 




fc 






t-l 




0> 


-Si 


> „; 


Sfl 


o 2 


V. 


o •— . 


*> 


<u <■<- 




38 AHLSTRAND 

fires in the parks have been limited to relatively small areas, having been con- 
tained by natural barriers or suppressed by fire crews. Successful fire sup- 
pression efforts and the elimination of livestock grazing in both parks have 
resulted in a gradual accumulation of fuel over the years. With the present 
fuel loads in many portions of the parks, an increase in the number, size, and 
intensity of fires can be expected. Successive cat-face scars on trunks of 
Juniperus deppeana throughout the back country of Carlsbad Caverns 
National Park suggest that extensive fires have occurred periodically in the 
past. 

VEGETATION ANALYSIS 

An attempt was made to examine all burns, one-tenth acre or larger in 
size, listed in the fire records of each park. It was not possible to identify the 
boundaries of burns that occurred prior to 1968. Of more than 35 burns 
located, eight were selected in the succulent desert formation for a compara- 
tive vegetation analysis. One of the eight selected for study was reburned 
during the summer of 1974 before sampling could be completed. 

For each burn, a 25 by 50 m permanent plot was established on the burn 
and on a nearly identical adjacent unburned stand. Cover was determined 
for shrubs by species along ten 25-m line intercepts placed at 5-m intervals 
within each plot. Species within a 20 by 50 cm plot frame placed at 5-m 
intervals along each 25-m line were recorded by species according to one of 
six cover classes (fifty 0. 1 m 2 subplots per 25 by 50 m plot). Intercept data 
were converted to coverage for individual species. Coverage and frequency 
values were calculated for species sampled in the subplots. Coverage values 
for species sampled in common by the two methods were in close agreement. 

Species shared in common by pairs of stands for a given burn averaged less 
than 50% of the combined total species for the pairs, which suggested the 
presence of serai species on the burned plots. However, when species pres- 
ent in all control stands were compared with species present in all burn 
stands, no serai species were indicated. When the same comparison was 
made after eliminating from consideration those species with frequency 
values of 2% or coverage of 1% or less, species shared in common between 
control and burn pairs averaged nearly 100%. Species area curves indicated 
that the stands were oversampled by a factor of 2 in terms of taxa present. 
However, plots of cumulative mean coverage showed that between 40 and 50 
subplots per stand were necessary to achieve coverage data. 

Most of the fires have not been of an intensity to eliminate many species. 
The competitive balance among species was altered by burning, but not to 
the extent that invasion of the site by serai species was permitted. The most 
evident effect of fire was to change the relative cover values for species 
present. Coverage of grasses and forbs often increased after burning, 
whereas that of shrubs usually was reduced (Table 5). Similar results were 



FIRE ECOLOGY 



39 



c 

H 

3 

c r- 



— i-H 

5 •"- 





4= 




00 


tA 


'S3 00 


*o 


c s© 

3 


C3 


a 






u 












o 








p 


o 

Is 


C3 


a 






SQ 








T3 




<D 
















3 


&• 


X) 


c^- 


,_ 


a> 


,o 


c 



CQ 



so 


«o 


P- 


o 


VD 


■^t 


1-1 


OO 


(N 


ON 


On 


SO 


ro 


SO 

! 1 


o 


On' 



OO 1—1 



-* 


co 


SO 


t 


1—1 


Tf' 


<N 


in 


°5 


On' 


CN 


t> 



^H (N 


T^ 


Tf 


CO 


r-H O 


os" 


cs 


<N 


CO 




"t 


00 


<N <N 


,-H 


(N 


r^ 


<N d 


o 


so 


so 


iO 




rH 


r- 









rD 




Cfl 




3 




32 












J3 




O 




on 




e 




>> 


| 


■g 


•2 


T3 

O 


o 


6 


o 


o 



in 


i-H 


— 1 


d 


OO* 


T3 


CO 




" 


<N 


SO 




«-H 


00 


(N 


O 


t-4 


£ 


** 


■«fr 


<N 


'St 


lO 


03 

C/3 





£ 


Tf 


0> 


OO 


£J 


so 






>> 


I> 


■O 


as 


bfl 


SO 


C 



f- 


CO 


SO 


<d 


"* 


CO 


CN 


a> 


«—i 


rt 


r- 








T3 


^H 


© 


SO 


O 

s 


SO 


SO 


On 






r» 


T3 
si 



o 

7j 



40 AHLSTRAND 

reported by Dwyer and Pieper (1967) for range land in south-central New 
Mexico. 

Responses of individual taxa to fire were not consistent. Most of the 

variation observed probably is due to differences in fire intensities and in soil 
moisture availability following the burn. Factual data concerning the role of 
fire in the Chihuahuan Desert region are difficult to obtain because this has 
been a topic of little research (Humphrey 1974). Limited observations on the 
reaction of individual taxa to fire in the Chihuahuan Desert were reported by 
Kittams (1972) and Humphrey (1974). Information concerning shrub 
species whose distribution includes the Chihuahuan Desert region has been 
reported in studies conducted elsewhere (Reynolds and Bohning 1956; Pond 
and Cable 1960; Dwyer and Pieper 1967; White 1969; Paseand Lindenmuth 
1971; Cable 1972; Wink and Wright 1973; Heirman and Wright 1973; 
Wright 1974a). Conflicting responses to fire reported for many taxa were at- 
tributed to varying seasonal and other environmental conditions that have 
seldom been well defined (Pieper et al. 1973). Wright (1974b) reported on the 
long-term effect of fire on several grasses common to this area and related 
the degree of response within individual taxa to annual precipitation. 

EXPERIMENTAL PRESCRIPTION BURNS 

A series of experimental burns is planned in order to document the 
response of vegetation to fires under different burning conditions and fuel 
types. The behavior of both headfires and backfires will be documented in 
several habitat types for five ranges of fuel moisture — 3 to 5%, 6 to 8%, 9 to 
11%, 12 to 14%, and 15 to 17%. Attempts will be made to burn at the 
midpoint intervals. Tentative plans call for burning when winds do not 
exceed 25 kmph, air temperature is between 10 and 30° C, and relative 
humidity is between 20 and 60%. 

The plot dimensions for each burn will be 10 by 10 m, with a 5-m buffer 
strip on each side to eliminate edge effects. Total area to be burned per fire, 
including the buffer strip, will be 400 m 2 . A fire line at least 1 m wide will be 
cleared completely around the outside edge of each buffer strip. Individual 
plots for each of the 10 burns per habitat type will be picked randomly from a 
grid of 12 plots, leaving 2 plots per grid as controls. 

A vegetation analysis of each plot will precede the actual burning and will 
be repeated periodically to follow recovery rates. Plant-water relations will 
be followed in selected taxa on burned and control plots. Fire intensity can 
be calculated as the product of the rate of fire spread and fuel energy con- 
sumed (van Wagtendonk 1974). 

SEED GERMINATION 

Studies of temperature and water potential limits for germination, and the 
effect of fire on viability and germination are needed for selected species for 
which seeds serve as the primary means of propagation. Information from 
these studies, when considered with energy budget data from various habi- 



FIRE ECOLOGY 41 

tats, can be used to estimate the success of different taxa on fire-altered 
habitats. 

Although no controlled environment studies of seed germination have 
commenced, a simple test was conducted twice in the laboratory with seeds 
harvested in November 1973. Seeds were placed in petri dishes on three 
layers of moistened filter paper. Tap water was added as needed. Light and 
temperature conditions were variable. Seeds were given no special prelreat- 
ments. Two replicates per trial were prepared and each replicate consisted of 
50 to 1 50 seeds, depending on seed size. Germination counts were made daily 
and the results are summarized in Table 6. 



TABLE 6. Results from two seed germination trials. Percent germination is on the fifteenth 
day of each trial. 







Germination (%) 


Average number 
of days to reach 












Species 


Feb- 


-Mar 


trial 


Aug-Sept trial 


50% germination 


Agave lechuguilla 




93 




88 


4 


Arbutus xalapensis 




80 




90 


10 


Bouteloua curtipendula 




3 




23 


- 


Bouteloua gracilis 




5 




39 


- 


Bouteloua hirsuta 




12 




13 


- 


Dasylirion leiophyllum 




9 




0a 


- 


Yucca torreyi 




98 




98 


7 



a Seeds infected with a fungus and failed to germinate. 

Greater germination percentages were attained by Bouteloua gracilis and 
B. curtipendula during the second trial and most likely were due to after- 
ripening. Future studies should investigate the effect of stratification on 
seeds of species slow to germinate or with low germination percentages. Seed 
viability for some species will be tested with 2,3,5-triphenyl tetrazolium 
chloride. 

Agave lecheguilla is a prolific seed producer and the seeds germinate 
readily in i elation to other species tested. Apparently, this is an important 
factor in its success on sites where competition by other species has been sub- 
stantially reduced, such as on overgrazed lands. 

Although seed viability in Arbutus xalapensis was high, growth of seed- 
lings progressed slowly in the laboratory. Field conditions necessary for 
germination of seeds and establishment of the species probably occur infre- 
quently. Even in those years favorable for germination, seedlings may die 
because their roots cannot keep pace with the descending depth of available 
soil moisture. 

PLANT WATER STATUS 

Important in the consideration of relict species, and especially those 
occupying tension zones, is their ability to survive in conditions imposed by 



42 AHLSTRAND 

fire-altered habitats. Water availability is undoubtedly the greatest single 
limiting factor controlling the distribution of plants in this area. Infor- 
mation concerning the seasonal water status of these species under present 
conditions is important in predicting their future success in modified micro- 
environments. 

Some plants are capable of regulating the magnitude of internal moisture 
stress, whereas others lack this ability and conform to the various stresses of 
their environment (Hickman 1970). The pressure chamber technique for 
measuring moisture relations in plants (Scholander et al. 1965; Waring and 
Cleary 1967) affords a convenient means for following the functional 
moisture stress in many plants under field conditions. Several taxa have been 
monitored periodically at Carlsbad Caverns National Park since October 
1974. Representative species and the range of water potentials observed with 
the pressure chamber were as follows: Pinus edulis, -14 to -17 bars; Juni- 
peruspinchotii, -17 to -20 bars; Berberis trifoliolata, -30 to -36 bars. These 
values represent baseline data, as none of the species was subjected to 
moisture stress during the monitoring period. As the drying season pro- 
gresses, various adaptations of different taxa to decreasing moisture supplies 
will be sought by following their functional water stress patterns with diurnal 
measurements. 

Additional studies concerning the moisture status of plants will include 
observations of stomatal diffusive resistances and cell osmotic potentials. 
Arid land plants possess numerous physiologic, morphologic, and anatomic 
adaptations for conserving moisture. Stomatal function is important to 
water economy. At finite transpiration rates, the lower limit of soil moisture 
availability is determined conceptually by the mesophyll osmotic potential, 
the resistance to water flow (controlled largely by the stomates), and the rate 
of transpiration. When the stomates are closed and the transpiration rate is 
zero, the cell osmotic potential determines the lower limit of soil moisture 
availability to the plant. Therefore, plants with low osmotic potentials have a 
definite advantage in moisture competition with plants possessing higher 
osmotic potentials, other factors being similar. Cell osmotic potentials and 
stomatal diffusive resistances will be observed for species during the inten- 
sive study of functional moisture stress. 

ENERGY BUDGETS 

The degree of microenvironment modification of fire-altered habitats 
varies according to the intensity of the burn. Where fuel loads are light, 
burning will effect little change. Some areas presently have fuel loads, which, 
if burned under the right set of environmental conditions, would lead to 
significant changes. Information concerning the degree of change that can be 
expected under given burning conditions in specific habitats is important in 
predicting which taxa can survive in the fire-altered habitats, and what long- 
term effects on soil, soil microorganisms, and faunal populations can be 
expected. The degree of change can be estimated for each community type by 



FIRE ECOLOGY 43 

a consideration of the energy budget for a two-dimensional surface (Lowry 
1969). 

No studies are planned on this phase of the investigation until more 
information from other parts of the study are available. 

LITERATURE CITED 

Agee, J. K. 1974. Fire management in the national parks. West. Wildlands 1:27-33. 

Cable, D. R. 1972. Fire effects in southwestern semidesert grass-shrub communi- 
ties. Proc. Tall Timbers Fire Ecol. Conf. 12:109-127. 

Cooper, C. F. 1960. Changes in vegetation, structure, and growth of southwestern 
pine forests since white settlement. Ecol. Monogr. 30:129-164. 

Dwyer, D. D., and R. D. Pieper. 1967. Fire effects on blue grama-pinyon-juniper 
rangeland in New Mexico. J. Range Manage. 20:359-362. 

Hanks, J. P., and W. A. Dick-Peddie. 1974. Vegetation patterns of the White 
Mountains, New Mexico. Southwest Nat. 18:371-381. 

Hastings, J. R., and R. M. Turner. 1965. The Changing Mile. Univ. of Arizona 
Press. Tucson, 317 pp. 

Heirman, A. L., and H. A.Wright. 1973. Fire in medium fuels of West Texas. /. 
Range Manage 26:331-335. 

Hickman, J. C. 1970. Seasonal course of xylem sap tension. Ecology 5:1052-1056. 

Humphrey, R. R. 1953. The desert grassland, past and present. J. Range Manage. 
6:159-164. 

1974. Fire in the deserts and desert grassland of North America. Pages 

365-400 in T. T. Kozlowski and C. E. Ahlgren, eds. Fire and Ecosystems, 
Academic Press, New York. 

Kittams, W. H. 1972. Effect of fire on vegetation of the Chihuahuan Desert region. 
Proc. Tall Timbers Fire Ecol. Conf. 12:427-444. 

Leopold, A. S., S. A. Cain, C. M. Cottam, I. M. Gabrielson, and T. L. 
Kimball. 1963. Wildlife management in the national parks. Trans. N. Am. Wildl. 
Nat. Resour. Conf. 28:1-18. 

Lowry, W. P. 1969. Weather and Life. An Introduction to Biometeorology. Aca- 
demic Press, New York, 305 pp. 

Pase,C. P., and A. W. Lindenmuth, Jr. 1971. Effects of prescribed fire on vegeta- 
tion and sediment in oak-mountain mahogany chaparral. /. For. 69:800-805. 

Pieper, R. D., D. D. Dwyer, and W. W. Wile. 1973. Burning and fertilizing blue 
grama range in South-Central New Mexico. N.M. Agric. Exp. Stn. Bull. 611:1-21. 

Pond, F. W., and D. R. Cable. 1960. Effect of heat treatment on sprout produc- 
tion of some shrubs of the chaparral in Central Arizona./. Range Manage. 13:313- 
317. 

POPE, J. 1854. Report of exploration of a route for the Pacific Railroad near the 
thirty-second parallel of north latitutde, from the Red River to the Rio Grande. 
Report to the War Department, 185 pp. (Copy available at Carlsbad Caverns and 
Guadalupe Mountains National Parks Headquarters Office, Carlsbad, New 
Mexico.) 

Reynolds, H. G., and J. W. Bohning. 1956. Effects of burning on a desert grass- 
shrub range in southern Arizona. Ecology 37:769 777. 

Robinson, J. L. 1969. Forest survey of Guadalupe Mountains, Texas. M.S. Thesis, 
Univ. New Mexico, Albuquerque, 71 pp. 



44 AHLSTRAND 

Scholander, P. R, H. T. Hammel, E. D. Bradstreet, and E. A. Hemming- 
SEN. 1965. Sap pressure in vascular plants. Science 148:339-346. 

Stone, E. C. 1965. Preserving vegetation in parks and wilderness. Science 150: 1261- 
1267. 

U.S. Department of the Interior, National Park Service. 1970. Compila- 
tion of the Administrative Policies for the National Parks and National 
Monuments of Scientific Significance (Natural Area Category). U.S. Government 
Printing Office, Washington, D.C., 147 pp. 

Van Wagtendonk, J. W. 1974. Refined Burning Prescriptions for Yosemite Na- 
tional Park, U.S. Dep. Inter. Occas. Pap. No. 2. U.S. Government Printing Of- 
fice, Washington, D.C., 46 pp. 

Waring, R. H.,andB. D.Cleary. 1967. Plant moisture stress: Evaluation by pres- 
sure bomb. Science 155:1248-1250. 

White, L. D. 1969. Effects of a wildfire on several desert grassland shrub species. J. 
Range Manage. 22:284-285. 

Wink, R. L., and H. A. Wright. 1973. Effects of fire on an ashe juniper commu- 
nity. J. Range Manage. 26:326-329. 

WRIGHT, H. A. 1974a. Range burning. J. Range Manage. 27:5-11. 

1974b. Effect of fire on southern mixed prairie grasses. J. Range Manage. 

27:417-419. 



ACKNOWLEDGMENTS 

I wish to acknowledge gratefully the assistance with the vegetation 
analysis rendered by former Park Technician Michael R. Glass. Sincere 
appreciation is extended to Staff Interpretive and Environmental Services 
Specialist Philip F. Van Cleave for sharing resource information pertinent to 
the study from his vast store of knowledge, and to Personnel Assistant 
Shirley A. Childress for proofreading and typing the manuscript. 

This study is supported by the National Park Service as part of its con- 
tinuing research program to provide information necessary for the manage- 
ment of the natural resources. 



The Guadalupe Mountains— A Chink 
in the Mosaic of the Chihuahuan 
Desert? 



MARSHALL C. JOHNSTON, University of Texas, Austin 

My purpose is to place the Guadalupe Mountains, botanically, into a 
broad regional framework, on the basis of floristic and vegetational 
evidence. This has relevance in the current trend toward studying the floras 
of biotically coherent geographic units instead of the traditional (and 
probably still more practical) political-geographic units. One of the first of 
the biotic units delineated as an area for special floristic study was the so- 
called Sonoran Desert, which Shreve (1942) courageously delineated and 
which Wiggins wrote up floristically (Shreve and Wiggins 1964). Shreve 
(1942) also attempted to delineate the Chihuahuan Desert, with which he 
was less well acquainted. 

In 1971, in proposing to write a Chihuahuan Desert Flora, I found it 
necessary and desirable to draw a line on the map to delineate the Chi- 
huahuan Desert Region. Anybody who has tried to draw a line on a map 
purporting to separate biotically distinct regions knows how hazardous this 
can be to the integrity of communication. For the line has an unwarranted 
narrowness and inflexibility that contradicts everything we as field 
observers know about biotic, climatic, and substrate continuity, about the 
usual gradualness of transitions, and about the mosaic nature of the sub- 
strate and of the climate in those areas of much topographic relief. All these 
conditions conspire to give us in nature both tightly chinked biotic mosaics 
and broad biotic transitions that defy our line-drawing attempts. As a poet 
once said, "Nature mocks at human categories." 

Nevertheless, lines are drawn on maps and in some cases they more or less 
successfully delineate biotic areas; in my case the line-drawing was a prac- 
tical necessity. Lines thus drawn always represent compromises, and mine 
was no exception. 

The line (Fig. 1) excludes the uppermost parts of the Guadalupe Moun- 
tains, which in fact constitute part of the northern boundary of the Chihua- 
huan Desert Region. Some confidence in this placement of the boundary 
may be inspired by the general consensus expressed in a recent symposium 
on the Chihuahuan Desert at Alpine, Texas. Many of the participants who 

45 



46 JOHNSTON 







106 


100 






| i 


D ' 


/— V 1 ' » V m 


i 










( ^ 








_ | 






1 ^ 






1 






1 
j 


"V " 






1 






1 
y ^-'GUADALUPE MOUNTAINS 








I \j^ 




\ \.^^C i 






32 




i— — \J — 


S^L _\j 




32 




\ 












"X 










\S Jyz 

\ \ 












j 

i 
) 
\ 




s ( / \ 








{ 




/ V ^ 






28 


V 

s 








28 














) ' l 


\ i 

/ 
\ 


L — V \f 7 ■ 

^ r v \ \ ^_ 


/ 








\ 


r" - 
f 


( N V \ 


/ 




24 




\ < 


f 


I r i ? > 


- 


24 






^ ) ) V 










\ I 


/ ( ^ y \ 












( \ \ ^° 1 










\ ( 
















^ 










\ 
















1 


i ] 


i ^r V i r , / 


y L 






106 


100 







FIG. 1. Sketch-map showing the extent of the Chihuahuan Desert Region and the 
placement of the Guadalupe Mountains. 



delineated the northern part of the chihuahuan Desert Region drew it in this 
area to exclude the higher part of the Guadalupe Mountains. 

Those of us who have worked a long time with the flora of western Texas 
and northern Mexico are impressed by the botanical features of the Guada- 



FLORAL RELATIONSHIPS 47 

lupe Mountains, which seem to us to reflect strong influence of the southern 
Rocky Mountains and the mounainous parts of the Colorado Plateau, par- 
ticularly those massifs which are largely calcareous. As examples, on the up- 
per slopes we find the common ground cover of Berberis repens Lindley with 
Fragaria bracteata Heller and Carexe burnea Boott here and there, and the 
poisonous Hymenoxys richardsonii (Hooker) Cockerell var. floribunda 
(Gray) K. Parker on the overgrazed exposures, and much Pinus edulis 
Engelm. mixed with the other gymnospermous plants. In the better-watered 
canyons we find such plants as Amelanchier utahensis K'oehne, Aster 
hesperius Gray, Lonicera arizonica Rehder, Physocarpus monogynus 
(Torrey) Coulter, Polygonatum cobrense (Wooton & Standley) Gates, 
Robinia neomexicana Gray, Valeriana arizonica Gray, and Zigadenus 
elegans Pursh, all of which are very special to us, reminiscent of the Rocky 
Mountains, and found nowhere else in Texas and nowhere in the Chi- 
huahuan Desert Region. 

As the only one of our Texas "front ranges" over 2000 m in altitude which 
is of calcareous rock (the Davis Mountains at 2500 m and the Chisos Moun- 
tains at 2400 m are of igneous rocks), the Guadalupe Mountains may be 
expected to show some other interesting, and, for Texas, unique floristic 
elements. Figure 2, in fact, shows what a truly massive and abrupt topo- 
graphic, climatic, and biotic barrier these mountains are. Thus it may not be 
too surprising to find here such widespread north-temperate species as 
Cystopteris bulbifera (Linnaeus) Bernhardi, Glyeeria striata (Lamarck) 
Hitchcock, Lactuca graminifolia Michaux, and Lilium philadelphicum 
Linnaeus var. andinum (Nuttall) Ker-Gawler, all of which are rather 
strongly mesophytic or even aquatic in their preferences, and all of which are 
rare or unknown elsewhere in Texas. 

Neither is it too surprising that a few species which are found in the lime- 
stone Edwards Plateau 200 km to the southeast have small disjunct popu- 
lations in the Guadalupe Mountains, which provide a relatively mesic island 
for their survival. These species include Lithospermum parksii I. M. 
Johnston var. rugulosum I. M. Johnston and Liatris punctata Hooker. 

All these botanical components have led us to emphasize the relation- 
ships of the Guadalupe Mountains to biotic regions other than the Chi- 
huahuan Desert Region, and I think it is wise and expedient to adhere to this 
exclusion. Thus I answer the question posed by my title in the negative, at 
least as it pertains to the uppermost parts of the Guadalupe Mountains. 

Before we become too settled in our convictions, however, I think it is 
desirable briefly to look at the Guadalupe Mountains from a more northerly 
vantage point, let us say from the point of view of a botanist working in the 
northern half of New Mexico. Such a botanist visiting the Guadalupe Moun- 
tains would be struck primarily by the extent that the truly montane 
vegetation is areally restricted and depauperate. A suggestion of affinity to 
the southern Rockies would strike him as a trifle far-fetched and strained, 
and he would be more inclined to emphasize the numerous species which 



48 JOHNSTON 



, 


T 


i< ■ ■ i i 


- 107° W. P 


Is 

)£, New Mexico - Chihuahua state line 


fc 


< 


\x 


< 

» Rio Grande between Texas and New Mexico 


i '"<* 






^_~a^fe Franklin Mountains, Texas 


ff°&*^~^ 


(^ "Jornada del Muerto" 


\>jj 


- 106° w. ^z^^eOn 


< ^>mz&L Hueco Mountains, Texas 


<0 


4r^ 


F 


*?= Cornudas Mountains, Texas 




^ 




J! Salt Flats, Texas 

ii 

c 




c 


- 105° W. ~& 


^<g^ Guadalupe Mountains, Texas 


- Turn ESE J^ 


)<. 


\<. 


fa 


S< 


i 


104° W.£ 


\>-l 






/^ 


[<Z 


/<c 


h> 


K 


X 


/V 


M Pecos River at Pecos, Texas 


1 


At 




< 




1 




< 




< 




c 


"Turn NE 


< 

'£j Pyote, Texas 




f 




U - r\) 


- 103° W. 


£ 9 o o 
< o o o 




H ., O 




< 3 3 




£ • • 3 




ff. . ; 



FLORAL RELATIONSHIPS 49 

find their northern, northeastern, or northwestern limits in or near these 
mountains, indicating more southern affinities. The floristic elements at 
moderate elevations which are at or near their northernmost distributional 
limits in the Guadalupe Mountains include Arbutus xalapensis Kunth, 
Chrysactinia mexicana Gray, Sophora secundiflora (Ortega) DeCandolle 
and its relatives, Juglans microcarpa Berlandier, Dichondra brachypoda 
Wooton & Standley, Quercus pungens Liebman, Rhamnus serrata Schultes, 
Selaginella lepidophylla (Hooker & Greville) Spring, Selaginella pilifera A. 
Brown, and Menodora longiflora Gray. At lower elevations more and more 
strongly xerophytic elements are found at or very near their northernmost or 
northwesternmost limits; these include Phyllanthus polygonoides Sprengel, 
Jatropha dioica Cavanilles, Bernardia obovata I. M. Johnston, Condalia 
warnockii M. C. Johnston, Leucophyllum minus Gray, Stenandrium 
barbatum Torrey & Gray, Ibervillea tenuisecta (Gray) Small, Ruellia parryi 
Gray, Mammillaria lasiacantha Engelmann, and a number of other Cac- 
taceae, Viguiera stenolaba Blake, Boerhavia linearifolia Gray, Mimosa 
biuncifera Bentham, Cassia lindheimeriana Scheele, and Acleisanthes longi- 
flora Gray. 

In summary, it appears that a balanced view of the Guadalupe Mountains 
will include a recognition of a dilute, weak, fragile, marginal, and areally 
restricted vegetational island reminiscent of the Southern Rockies and the 
mountains of the Colorado Plateau, as well as a vast area surrounding on 
three sides which is clearly part of the Chihuahuan Desert Region. In fact, 
there are mountain ranges wholly within the Chihuahuan Desert Region 
that strongly resemble the Guadalupe Mountains and that support more 
extensive gymnospermous forests, e.g., the Sierra de la Madera, Coahuila. 
The reason they are included within the Chihuahuan Desert Region is that 
they are more centrally located and completely surrounded by desert vegeta- 
tion. The Guadalupe Mountains, at least the upper 500 m, are excluded 
because they are surrounded on only three sides by desert, and they are at the 
northern limit of that desert vegetation. 

LITERATURE CITED 

Shreve, F. 1942. The desert vegetation of North America. Bot. Rev. 8:195-246. 
Shreve, F., and I. L. Wiggins. 1964. Vegetation and Flora of the Sonoran Desert. 
Stanford Univ. Press, Stanford, California (2 vols.), 1740 pp. 



Summary of the Vegetative Zones 
of the Guadalupe Mountains 
National Park, Texas 



DAVID K. NORTHINGTON and TONY L. BURGESS, 
Texas Tech University, Lubbock 

As even the nonbiologist could readily notice, the vegetation of the 
Guadalupe Mountains National Park is extremely varied and often of 
unique composition. Topographic, climatic, and edaphic factors interact to 
produce this vegetational mosaic in which the delimitation of general zones 
is often difficult if not impossible. Our purpose is to summarize the various 
vegetative regions of the Guadalupe Mountains National Park, with special 
attention given to those transitional and unique associations that defy such 
classification. 

Warnock (undated) and Gehlbach (1967) have treated the vegetative com- 
munities of the Guadalupe Mountains region, thus some of our observa- 
tions represent duplication. Gehlbach, however, described only the eastern 
escarpment, primarily McKittrick Canyon, and neither author commented 
adequately on certain portions of the park's higher elevations. Table 1 sum- 
marizes briefly the vegetative zones and specific plant associations of 
Warnock and Gehlbach. Potter and Robinson (1968) did describe a con- 
siderable portion of the coniferous forest areas in the park; again, however, 
some areas of interest were not included in their study area. In addition to 
these references, Glass and Reisch ( 1974) conducted an Interagency Browse 
Survey for the Guadalupe Mountains National Park and included therein is 
a detailed vegetation map for the entire park. The plant names used in this 
paper correspond to those in Correll and Johnston (1970). 

VEGETATIVE ZONES 

For sake of simplicity, and in many cases for accuracy, we will categorize 
the park into three general biotic communities — desert, woodland, and 
forest. Specific plant associations within each of these zones can then be 
treated individually, whether a specific descriptive tag can be applied or not. 

51 



52 



NORTHINGTON AND BURGESS 



TABLE 1. Two schemes of vegetation classification for the Guadalupe Mountains. B. H. 
Warnock's (undated) formations and associations are in the left column. F. R. Gehlbach's 
( 1 967) system for the east escarpment only is listed by formations and dominance types in the 
right column. Nomenclature changes have been made to conform species names to Correll 
and Johnston (1970). 



WARNOCK 



GEHLBACH 



Desertscrub 

Pro sop is glandulosa 

Larrea tridentata-Prosopis glandulosa 

Grassland 

Agave lecheguilla- Dasylirion 

leiophyllum 
Bouteloua eriopoda-Aristida 

pansa 
Dasylirion leiophyllum - Forsellesia 

spine scens 
Dasylirion leiophyllum - Cercocarpus 

montanus 

Woodland 

Chaparral: Cercocarpus montanus- 
Ceanothus greggii- Que reus 
undulata-Q. pungens 

Evergreen/ deciduous complex: 
Quercus sp.-Juniperus sp.- 
Pinus edulis-Mimosa borealis- 
Acacia neovernicosa 

Canyon: Arbutus xalapensis-Acer 
grandidentatum-Prunus serotina- 
Quercus muhlenbergii-Q. grisea- 
Juniperus monosperma-J. deppeana- 
Pinus edulis 

Riparian: Populus sp.-Salix sp.- 
Juglans microcarpa- Celt is 
reticulata 

Coniferous forest 
Pinus ponderosa 
Pseudotsuga menziesii- Pinus 
strobiformis-P. ponderosa 



Shrub desert 

Larrea tridentata- Flour ensia cernua 

Acacia neovernicosa- Larrea tridentata 

Succulent desert 

Dasylirion leiophyllum- Agave 

lecheguilla 
Juniperus monosperma- Agave 

lecheguilla 



Evergreen woodland 
Quercus grisea- Juniperus 

(monosperma? ft 
Quercus grisea- Juniperus deppeana 
Juniperus deppeana- Pinus edulis 



Deciduous woodland 

Juglans microcarpa- Dasylirion 

leiophyllum 
Fraxinus velutina-Ostrya knowltonii 

Coniferous forest 
Pinus ponderosa 
Pseudotsuga menziesii- Pinus 
strobiformis 



Desert 

Considerable portions of the region below the west and east escarpments 
would qualify as desert. Because of the dominance of microphyllous shrubs 
such as Larrea tridentata, Prosopis glandulosa, Acacia neovernicosa, 
Flourensia cernua, and A triplex canescens, this vegetation is often referred 
to as desertscrub. On most of the west bajada below 5000 ft there is "typical" 



VEGETATIVE ZONES 53 

desertscrub in which Larrea dominates either totally or in association with 
Bouteloua eriopoda, Sporo bolus airoides, and Erioneuron pulchellum. 
Some of the more unique plant associations in the park occur in particular 
edaphic situations on the lower west side. A more thorough review of the 
desert vegetation of this region is available (Burgess and Northington 1975); 
however, a brief overview of some of the more salient features is presented 
herein. 

Near the edges of the Salt Flat, Sporobolus airoides dominates, with 
A triplex canescens, Frankenia jamesii, and Allenrolfea oecidentalis 
constituting an open shrub stratum. Heavy clay soils, for example, in the 
vicinity of Homsley's Dugout, support extensive stands of Atriplex 
canescens. Adjacent to this area, near the west boundary of the park, are 
gypsum dunes and ridges in various stages of stabilization. A mosaic of asso- 
ciations covers these sites dominated variously by Coldenia hispidissima, 
Bouteloua breviseta, Sporobolus nealleyi, Poliomintha incana, Ephedra 
torreyana, Yucca elata, and Opuntia polyacantha. 

To the northeast of this gypsum complex a large area of quartz sand 
occurs. These "red dunes" have Prosopis glandulosa, Atriplex canescens, 
Croton dioicus, and Yucca elata sharing dominance with a Sporobolus 
complex including S. giganteus, contractus, andflexuosus. In portions of 
these quartz sands, the grass species so dominate that the recognition of a 
grassland biotic community could be defended. In fact, Warnock would con- 
tend that most of the lower elevations (and often up to 7500 ft) of the park 
should be classified as grassland from which overgrazing has produced the 
current desertscrub communities. Such a view could well be supported in 
light of the geographic proximity of the plains grasslands immediately to the 
north and east. As Lowe (1964) pointed out, although the Chihuahuan 
Desert has more grasses than the Sonoran Desert and even in light of some 
investigators' contentions that the entire Chihuahuan Desert should be con- 
sidered a grassland climax, the current status of the region is that of a natural 
desert dominated by numerous climax shrubs. 

As previously pointed out, however, the position of the Guadalupe Moun- 
tains National Park is transitional between the Chihuahuan Desert and the 
plains grasslands. There are certainly clear-cut examples of both commu- 
nities within the park boundaries, but equally as prevalent are transitions in 
which a plethora of plant taxa are found variously in association. 

Whether the lower elevations of this region are disclimax grasslands 
invaded by succulents and microphyllous shrubs due to overgrazing, or a 
true desert climax with increased coverage of grass species due to the 
proximity of the plains grassland biome is certainly an interesting question. 
The fact that grassland associations of varying complexities do exist within 
the park cannot be ignored. Due to the great variability exhibited in relation 
to grass coverage, associated species presence, and edaphic influences, we 
feel it best to examine several major grassland communities within the park 
and treat exercises in classification as a futile game of semantics. 



54 NORTH INGTON AND BURGESS 

Some of the nicest examples of grassland associations occur just above 
the alluvial fans among the limestone slopes on the west escarpment. Such a 
community on the rocky slopes above Williams Ranch is dominated by 
Bouteloua eriopoda with significant numbers of Agave lecheguilla, Viguiera 
stenoloba, and Fouquieria splendens. This area certainly would be called 
grassland by Warnock, whereas Gehlbach would consider this as succulent 
desert. An alluvial fan in the northwest corner of the park reveals an open 
grass cover of Bouteloua eriopoda, Aristida glauca, and A. pansa, punc- 
tuated with scattered Yucca torreyi and Krameria glandulosa. 

Between 5500 and 6000 ft on the west escarpment, there exist areas of less 
slope and deeper soils derived from eroded sandstone which are almost 
totally grass dominated. The major grass taxa in these areas include 
Bouteloua hirsuta, B. warnockii, B. gracilis, B. eriopoda, Lycurus 
phleoides, Eragrostis lugens, Muhlenbergia setifolia, Stipa neomexicana, 
and Aristida glauca. In rockier areas these grass species become "invaded" 
by several shrub species, especially Dasylirion leiophyllum, Parthenium 
incanum, Viguiera stenoloba, Mortonia scabrella, Leucophyllum minus, 
and Daleaformosa. At higher elevations (above 6500 ft) on the west escarp- 
ment, dominance shifts from grass species such as Muhlenbergia pauci- 
flora, to Cercocarpus montanus, Choisya dumosa, Forsellesia spinosa, 
Nolina micrantha, Yucca baccata, and Dasylirion leiophyllum. 

On the more gradual slopes of the east escarpment are found grass- 
dominated areas which often have Dasylirion and Opuntia species mixed 
with Juniperus pinchotii, or in more riparian sites J. deppeana, Quercus 
grisea, and Arbutus xalapensis. The most common grass species in these 
grassland-succulent desert-woodland communities are Bouteloua curti- 
pendula, B. gracilis, Lycurus phleoides, Muhlenbergia setifolia, Aristida 
glauca, and Stipa neomexicana. Areas near Pine Spring Canyon and Nipple 
Hill show plant associations ranging from short-grass grassland to succu- 
lent desert-grassland to succulent desert-open woodland-grassland. Such 
complex mosaics render useless general classification schemes that would 
effectively delineate biome formations on a local basis. 

Possibly, the creation of intermediate terminology to describe the con- 
siderable acreage within the park that, depending on the author, could be 
classified as either grassland or succulent desert would be effective. To that 
end we will refer to these vegetative assemblages as "succulent grassland." 
Another possible term to describe this complex is "Ensotal" (M. C. 
Johnston, pers. comm.). 

Woodland 

As we have already pointed out, especially on the east escarpment, open 
woodland-grassland transitions occur, leading to many of the true wood- 
land associations. These transitions occur as one goes from the desert floor 
up into the more mesic and cooler arroyos and canyons. Warnock would 



VEGETATIVE ZONES 55 

refer to this association as canyon woodland. The more xeric of these 
riparian communities contain Juglans microcarpa and Dasylirion 
leiophyllum. On a xeric to mesic gradient, Quercus grisea (and some 
Juniperus monosperma) replace Juglans. These taxa are then replaced by 
Quercus undulata and Juniperus deppeana which grade into Quercus 
muhlenbergii, Ostrya knowltonii, and Acer grandidentatum. An accurate 
delimitation of deciduous as opposed to evergreen woodlands in these 
canyons is sometimes difficult although deciduous taxa usually 
predominate. 

Another factor that further complicates canyon vegetation structure is 
slope effect. As these riparian associations move out onto rocky slopes which 
are especially common in the McKittrick Canyon complex and on the west 
side, the nature of the vegetative structure changes yet again. The presence of 
low growing oaks, Dasylirion leiophyllum, Fouquieria splendens, 
Ceanothus greggii, Cercocarpus montanus, Mimosa biuncifera, and 
scattered Agave neomexicana present a chaparral-like association. Such 
associations also occur in similar habitats of Dog Canyon with the regular 
addition of Nolina micrantha. 

In addition to these riparian (deciduous), chaparral, and evergreen wood- 
land associations, one of the most unique vegetative zones in the entire park 
occurs in several localities of the McKittrick Canyon system. These areas 
represent a serai stage in the development of a deciduous woodland and are 
characterized by Adiantum capillis-veneris and later, Cladium jamacense . 
These mats or "hanging terraces" are potentially threatened by human visita- 
tion because relatively little trampling is sufficient to disturb them signifi- 
cantly. As these present a favorable habitat for distributional limits of sev- 
eral taxa (such as Cladium) in the region, concern for their preservation is 
well warranted. 

The lower slopes of both Upper Dog and West Dog Canyons are charac- 
terized by deeper and less rocky soils. These slopes have rather extensive 
stands of Pinus edulis and Juniperus monosperma sharing dominance (a 
Pinyon-Juniper evergreen woodland). In the alluvial bottoms are found 
grassy meadows of Stipa tenuissima or in some of the more xeric areas of 
West Dog Canyon, Bouteloua gracilis. 

Between Cox Tank and Coyote Peak occurs a unique region previously 
undescribed as a significant plant association in the Guadalupes. In scattered 
stands throughout this area Juniperus monosperma and Sophora 
gypsophila var. guadalupensis share co-dominance. Yucca baccata and 
Cercocarpus montanus are commonly associated species in these areas as are 
Bouteloua curtipendula and Aristida glauca in rockier regions. Also com- 
mon to the more xeric, steep, rocky slopes in this northwest canyon com- 
plex are Nolina micrantha, Dasylirion leiophyllum, and Agave neo- 
mexicana. As one moves along a xeric to mesic gradient, Juniperus mono- 
sperma is replaced by Pinus edulis and finally merges into the edges of coni- 
ferous forest forms. 



56 NORTHINGTON AND BURGESS 

Forest 

As would be expected by now, there exist numerous transitional areas 
between evergreen and/ or deciduous woodland zones and the easily recog- 
nizable coniferous forest regions. In the more xeric forest associations, Pinus 
ponderosa dominates with significant amounts of Muhlenbergia pauciflora 
(and Quercus gambelii) on open slopes. The more mesic forest associations 
have Pseudotsuga menziesii and Pinus strobiformis sharing dominance with 
Pinus ponderosa as an associated subdominant in some areas. The most 
typical coniferous forest associations occur in The Bowl, but notable stands 
occur throughout the higher elevations near Bush Mountain, Blue Ridge, 
and the McKittrick Canyon drainage. One small stand of Populus tremu- 
loides still occurs west of The Bowl where it is locally still a subdominant. 

Rock Outcrops 

A plant association that warrants attention in this overview but does not 
fit naturally in any of the preceding three general vegetation zones is the 
complex of taxa that inhabits the steep limestone rock faces throughout the 
park. Although many of these habitats do occur in association with the 
riparian woodlands of the McKittrick drainage system, they also occur on 
the east and west escarpment faces and in portions of the Dog Canyon 
drainage in association with everything from almost desert to montane for- 
mations. In these habitats are found a large number of the unique taxa from 
the park including Chaetopappa hersheyi, Cystopteris bulbifera, and C. 
fragilis, Epithelantha micromeris, Hedeoma apiculatum, Nama xylopodum, 
Perityle quinqueflora, Phanerophlebia auriculata, Pinaropappus parvus, 
and Salvia summa. In addition to these "rock endemics," Fendlerella 
utahensis, Fendlera rupicola, Petrophytum caespitosum, Hedyotis 
nigricans, and Philadelphus hitchcockianus are also found in these habitats. 
Thus each of these areas is a separate and fairly distinct group of associa- 
tions that is of particular interest due to the numbers of taxa restricted to 
such habitats which are in some way unique distributionally. 

SUMMARY 

Our purpose was to present a brief summary of the major vegetative zones 
in the Guadalupe Mountains National Park region with special emphasis 
given to selected plant associations that exhibit "typical" plant class forma- 
tions and to those that represent transitions between them. A thorough 
(quantitative) analysis of the vegetational associations in the Guadalupes is 
certainly needed as can be seen by the categorization difficulties encountered 
in this short overview. The critical message here is an awareness of the 
uniquely complex vegetational mosaics of this region produced by sudden 
and extreme topographic and edaphic interfaces in an essentially arid 
climate. These various floristic elements occur at a crossroads of major 
biotic assemblages: Rocky Mountain Forest; Chihuahuan Desert Scrub; 



VEGETATIVE ZONES 57 

Great Plains Grassland; and some elements of the Sierra Madrean Wood- 
land (Southwestern Mountains). This geographical position is in a zone of 
climatic interface which results in temporally unstable habitats containing 
unique plant associations. Such complexity is what makes this area so 
striking and interesting to both the scientist and to the general public. 
Because most of the area in question is part of the Guadalupe Mountains 
National Park, preservation of these features is more assured as is the oppor- 
tunity of exposing the public to nature at its heterogeneous best. 

LITERATURE CITED 

Burgess, T. L., and D. K. Northington. 1977. Desert vegetation in the Guada- 
lupe Mountains region. In R. H. Wauer and D. H. Riskind, eds. Transactions- 
Symposium on the Biological Resources of the Chihuahuan Desert Region, U.S. 
and Mexico, National Park Service, Washington, D.C., in press. 

Correll, D. S., and M C. Johnston. 1970. Manual of the Vascular Plants of 
Texas. Texas Res. Found., Renner, Texas, 1881 pp. 

Gehlbach, F. R. 1967. Vegetation of the Guadalupe Escarpment, New Mexico- 
Texas. Ecology 48:404-419. 

Glass, M., and R. Reisch. 1974. Summary of range conditions. Interagency 
Browse Analysis Survey, Guadalupe Mountains National Park, 1973, 16 pp. 

Lowe, C. H. 1964. The Vertebrates of Arizona. Univ. Arizon Press, Tucson, 270 
pp. 

Potter, L. D., and J. L. Robinson. 1968. Effects of development and use on the 
relict forest and woodland of the southern Guadalupe Mountains. Final report 
submitted to the National Park Service, U.S. Dept. of the Interior, Santa Fe, 
New Mexico. 

Warnock, B. H. Undated. Plant communities of the Guadalupe Mountains in 
Texas and nearby Carlsbad Caverns National Park. Report submitted to the 
Carlsbad Caverns National Park. 



ACKNOWLEDGMENTS 

We acknowledge gratefully the staff of the Guadalupe Mountains 
National Park for their cooperation and assistance with the logistical por- 
tions of our studies. Field work was supported by the Department of the 
Interior, National Park Service Grant Number CX-700040145. 



Status of Rare and Endangered Plant 
Species of the Guadalupe Mountains 
National Park, Texas 



DAVID K. NORTHINGTON and TONY L. BURGESS, 
Texas Tech University, Lubbock 

The flora of the Guadalupe Mountains National Park has been described 
accurately in general terms by almost every scientist who has visited the 
region. A summation of these observations would include comments about 
vegetational mosaics; community diversity with sudden transitions; and cer- 
tainly there would be reports of rare, endemic, or unique taxa occurring in 
the delicate habitats produced by such topographic and edaphic diversity. 
There is good cause for such profuse descriptive phraseology and equally 
good cause for the more recently escalating number of warnings and 
expressed fears for the future of this area. We certainly concur with all such 
descriptions and unfortunately we must join the growing numbers of people 
who feel a real and immediate concern for the future of this beautiful and 
unique region. This concern is based on our recent survey work in the park 
for the National Park Service. Although this study is still active and not com- 
plete, we feel that sufficient data are already available for inclusion in this 
symposium volume. 

This report will emphasize the unique taxa within the park and, based on 
these data, recommendations will be presented for management of the 
critical areas of the park. Appendix I summarizes most of the plant taxa that 
we consider unique, along with the criteria used to establish their degree of 
uniqueness. This list also includes distributional, ecological, and interpreta- 
tive comments where applicable. At a later date this list will be expanded and 
corrected where necessary as a result of continuing, but as yet unfinished, 
identification efforts by both our own and outside personnel. For the most 
part our taxonomy follows that of Correll and Johnston (1970). 

UNIQUE FLORISTIC FEATURES: AREA BREAKDOWN 

A brief presentation of notable floristic elements in discrete areas within 
the park follows. More information on each species is presented in Appen- 
dix I. 

59 



60 NORTHINGTON AND BURGESS 

McKittrick Drainage Complex 

The integrity of the McKittrick drainage is central to preservation of the 
unique flora of the park. It is the most mesic and extensive system within the 
range, and supports representative populations of a majority of the rare and 
endemic taxa. 

General 

For control of erosion and floods, maintenance of vegetative cover 
throughout the watershed is essential. This should not preclude the possi- 
bility of controlled burning to thin stands of young conifers at higher eleva- 
tions; in fact, by reducing available fuel and thereby lessening the intensity of 
accidental burns, this practice could be beneficial to the drainage system. We 
expect that Dr. G. M. Ahlstrand's research will provide needed information 
for burn management. 

Several endemic species can be found in crevices of limestone cliffs and 
ledges throughout this drainage. Among these are Nama xylopodum, 
Chaetopappa hersheyi, Pinaropappus parvus, Salvia summa, Hedeoma 
apiculatum, Valeriana texana, and Polygala rimulicola. Their preservation 
is aided by the inaccessible locations of some colonies. 

Forested canyon bottoms above about 7000 ft shelter disjunct popula- 
tions of several Rocky Mountain species, some found nowhere else in Texas. 
These include Fragaria bracteata, Swertia radiata, and Physocarpus 
monogynus. 

Many unique taxa are concentrated along streams and adjacent boulders 
and gravel alluvium. The accessibility and relatively small area of this 
habitat make it one of the most vulnerable in the park to human impact. 
Species in this habitat include Aquilegia chaplinei, Glyceria striata, 
Stephanomeria wrightii, Sisyrinchium demissum, Lactuca graminifolia, 
Rosa woodsii, Streptanthus sparsiflorus, Equisetum kansanum, Aster 
hesperius, Penstemon cardinalis, Asclepias tuberosa, and Viola 
missouriensis. 

Special Considerations 

South McKittrick. — A small canyon draining northeast from the earth tank 
in The Bowl appears more mesic than most canyons in the drainage, and it 
contains a well-developed forest flora including Smilacina racemosa, 
Corallorhiza striata, and Rosa woodsii. This canyon merits attention as a 
preservation area. 

The mats of vegetation which have developed on seeps in limestone ledges 
support several rare or endemic species {Aquilegia chaplinei, Zigadenus ele- 
gans, and Physocarpus monogynus). Often dominated by Cladium jamai- 
cense or Adiantum capillus-veneris, these are among the most unique plant 
associations in the Guadalupes. Within the park the best-developed mats or 
"hanging terraces" are located between Turtle Rock and The Narrows. The 
wet limestone base makes these mats unstable and vulnerable to foot traffic. 



STATUS OF PLAN IS 61 

The Narrows is an area of steep limestone cliffs and pools which obstruct 
passage up the narrow canyon bottom. Its location is approximately 2.6 km 
S, 2.0 km W Pratt Lodge. The ferns Cystopteris bulbifera and 
Phanerophlebia auriculata grow in shaded crevices, and Viola missouriensis 
is common on more sunlit sites. Protection of this area is aided by its remote- 
ness from existing trails. 

North McKittrick. Devil's Den contains most of the Celastrus scandens 
found in the park. A small colony of this species occurs on a slope and adja- 
cent streambed a short distance below a dry waterfall in the lower part of the 
canyon (about 0.5 km N, 1 . 1 km W Pratt Lodge). This area also supported 
the greatest density of Streptanthus sparsiflorus observed in 1974. 

The New Mexico segment of the south fork of North McKittrick contains 
an area of pools, seeps, and well-developed "hanging terraces." Most species 
of the unique canyon flora can be found here, including the only Lilium 
philadelphicum we could find in the Guadalupes. A few small Celastrus 
scandens occur near the junction of Devil's Den with main North McKittrick 
and a small trail into the area has become more evident during the past 2 
years, indicating increased usage. 

Lower McKittrick. — Widely scattered plants of Yucca faxoniana occur on 
slopes and small side canyons, and Sophora secundiflora also can be found 
in similar habitats. Unusual forms of Agave are scattered on alluvial terraces 
and slopes along with Agave neomexicana and A. lecheguilla. Grindelia 
havardii grows along the streambed. 

East Escarpment— General 

Lower elevations are a mixture of desert and grassland species, with cor- 
ridors of riparian woodland along the streambeds. There is a gradual transi- 
tion to coniferous forest at the top. Protection from overgrazing 
undoubtedly has favored growth of grasses and associated species. As in the 
McKittrick drainage, limestone cliffs of the escarpment support a flora with 
a high percentage of endemic species. 

Special Considerations 

Smith Spring supports small populations of Glyceria striata and Viola 
missouriensis, typical of McKittrick Canyon. In addition, Smith Canyon 
contains a pocket of forest where Heterotheca viscida, Asclepias tuberosa, 
Streptanthus sparsiflorus, and Penstemon cardinalis subsp. regalis have 
been collected. The area is easily visited by a relatively short walk. 

Both Choza Spring and Upper Pine Spring have communities of stream- 
side plants including Lobelia cardinalis. Manzanita Spring, though some- 
what trampled along the edges, also has a few stream-side species and con- 
tains the only Potamogeton found in the park. All three areas are of easy 
access. 

Yucca faxoniana is rare over most of the area; however, an aggregation of 
about 30 plants is located on the ridgetop south of Smith Spring, and a 
smaller group occurs on the north side of the mouth of Pine Spring Canyon. 



62 



NORTHINGTON AND BURGESS 



Upper Dog Canyon 

Riparian woodland along the streambed and a more xeric vegetation of 
sotol, grasses, and low shrubs on the slopes produce a diverse flora. Several 
trees of the undescribed drooping form of Juniperus scopulorum are found 
along the streambed about 300 m south of the ranger station. Lesquerella 
valida has been collected from several slopes in this area. 

Humphrey Canyon— PX Flat 

This area contains the most extensive Pinyon-Juniper woodland in the 
park. The only Sophora gypsophila found in the United States occur here 
(Fig. 1). 



> J- 





tlA< 



^\-y 




Fig. 1 . Heavy solid lines outline the distribution of Sophora gypsophila var. guada- 
lupensis in the Guadalupe Mountains National Park, Texas, as provided by Roger 
Reisch and field-checked by the authors; light, solid lines indicate contour; dot-dash 
lines represent intermittent streambeds; parallel dashed lines indicate a jeep road. 



West Side— General 

The West Side contains associations representative of most types of 
northern Chihuahuan Desert vegetation. Most of the bajada is covered by 
Larrea tridentata or a desert grassland dominated by Bouteloua eriopoda. 



STATUS OF PLANTS 63 

Coryphantha dasyacantha, Penstemon dasyphyllus, and Bouteloua 
warnockii have been collected on the slopes of the west escarpment, and 
most of the endemic limestone cliff plants have been recorded from the 
canyons. 

Special Considerations 

A small colony of Jatropha dioica occurs on the south slope of the south 
Stagecoach Hill (0.8 km N, 3.5 km E Lewis Well). 

Bone Spring is the only known location for Astragalus humistratus in the 
park, and supports a few plants of Lobelia cardinalis and Forestiera 
pubescens. A small group of Yucca faxoniana grows on the south rim of 
Bone Canyon. 

The lower elevations of the northwest corner of the park contain desert 
grassland where Opuntia schottii was collected. Lower canyons in this area 
have Sophora secundiflora along the streambeds. 

Areas of wind-deposited sand contain species not found elsewhere in the 
park, among them Penstemon ambiguus, Caesalpinia jamesii, Dalea 
scoparia, Oryzopsis hymenoides, Panicum ramisetum, and Sporo bolus 
giganteus. Gypsum outcrops and the few gypsum dunes included within the 
park support a flora with several species of restricted distribution, including 
Senecio warnockii, Gaillardia multiceps, Dicranocarpus parviflorus, 
Coryphantha scheeri, Nama carnosum, and Mentzelia humilis. The best 
example of a gypsum outcrop association is located about 1.2 km S, 5.5 km 
W summit Bush Mountain. The largest gypsum dune seen within the park 
is about 400 m southwest of Lewis Well; the most extensive gypsum dunes 
are immediately west of the park. 

PROPOSED MANAGEMENT FOR FLORISTIC PRESERVATION 

Ideally, there should be no compromises in the preservation of rare or 
endemic species. Unfortunately, limited resources and manpower neces- 
sitate alternatives. Policymakers should bear in mind that biological com- 
munities are not static, especially in the variable climate of the Trans-Pecos. 
Continued monitoring of resources and a certain amount of managerial 
flexibility permitting more stringent protection measures in dry years or 
during peak visitation periods are essential to adequate floristic preserva- 
tion. Managers are always faced with compromising recreational or 
preservational usage, and the acceptable level of community degredation 
appears to be largely a result of local policy. The staff of Guadalupe Moun- 
tains National Park is to be commended for its close monitoring of the area 
and its concern for preservation as reflected by management policies. To a 
arge extent, these policies appear to have been successful in minimizing 
damage in critical areas, and many of our recommendations coincide with 
current management practices. 

A hierarchy of the relative status of locations and species requiring preser- 
ational action follows. Criteria for ranking include: (1) degree of unique- 



64 NORTH1NGTON AND BURGESS 

ness (the number and kinds of species present); (2) size of the habitat; and (3) 
vulnerability to human impact — including substrate stability and current 
accessibility. We know of no objective function to integrate the above fac- 
tors, and the order is based largely on our personal judgment. The status of 
each area will require re-evaluation as new or improved roads and trails 
change visitor use patterns. 

Most Critical Areas 

South McKittrick Canyon between Turtle Rock and The Narrows. 
South fork of North McKittrick Canyon (including New Mexico seg- 
ment). 

Critical Areas 

Remainder of South McKittrick drainage, including higher areas between 

Bush Mountain and The Bowl. 
North McKittrick Canyon. 
Devil's Den. 

Forest and meadows on Blue Ridge. 
The Bowl. 
Smith Spring. 
Sophora gypsophila in West Dog Canyon. 

Require Periodic Monitoring for Possible Action 

Limestone cliff associations at higher elevations throughout the park. 

Jatropha dioica colony on south Stagecoach Hill. 

Upper Pine Spring. 

Choza Spring. 

Drooping form of Juniperus scopulorum in Upper Dog Canyon. 

Lesquerella valida in Upper Dog Canyon. 

Yucca faxoniana (Lower McKittrick Canyon, ridgetop south of Smith 

Spring, north side of mouth of Pine Spring Canyon, south rim of Bone 

Canyon). 
Populus tremuloides west of The Bowl. 
Bone Spring. 

Gypsum outcrop, 1.2 km S, 5.5 km W summit Bush Mountain. 
Gypsum sand and beach ridges west of Lewis Well. 
Quartz sandhills ("red sands") west of Bush Mountain. 

Recommendations 

The following management recommendations are arranged by area. 
Within each area they are ordered from (A) most desirable (minimal 
degradation from human impact) to (B) or (C) minimum required for 
resource preservation. Unless contradictory, assume measures in (B) or (C) 
are included in (A). The requirements and ecology of most unique species are 
largely unknown, and in some cases these alternatives are little more than 



STATUS OF PLANTS 65 

educated guesses. For proper management there is no substitute for con- 
tinued monitoring and informed revision of policy as necessary. 

Entire Park 

Grazing and Browsing. — (A) Monitor browsing and grazing impact con- 
tinuously and take appropriate measures to prevent extensive degrada- 
tion; to include thinning elk and deer populations. (B) Maintain fences to 
prevent trespass grazing by stock. 

Burning. — Policy should be based on the results of Dr. G. M. Ahlstrand's 
current research. Consideration should be given to protecting the 
drooping Juniperus scopulorum in Dog Canyon, the Yucca fa xoniana 
aggregations, and Sophora gypsophila from intense fires. 

Human traffic. — Specific recommendations are made for each area. In this 
report, we recognize the following categories of access: Controlled 
access — visitor entry monitored and if necessary limited; Limited access — 
daily visitor quota to distribute impact; and Restricted access — no entry 
except to interested investigators or small groups accompanied by a 
ranger interpreter. To maintain the integrity of the range no vehicular 
traffic should be allowed above 6000 ft elevation. The resulting increased 
impact on higher parts of the watershed would result in an unacceptable 
level of habitat degradation. 

McKittrick Drainage 

Lower Canyons. — (A) Permanent ranger stationed at Pratt Lodge; daily 
patrols. (B) Low visitor quota; irregular patrolling. (C) Limited access; 
day use only; moderate visitor quota; intensive patrolling. 

South McKittrick. — (A) Restricted access to area northeast of The Bowl and 
Guadalupe Trail and southwest of McKittrick Trail; canyons draining 
Blue Ridge — Bush Mountain area. (B) Restricted access to canyon bot- 
tom including The Narrows to south of Turtle Rock. Day use only for 
remainder of areas in (A). Visitor quota for high elevation camps. No 
overnight horse or mule use. Monitoring and periodic "resting" of 
camping areas. No new trails. 

North McKittrick.— (A) Restricted access to Devil's Den; south fork of 
North McKittrick, including New Mexico segment. (B) Day use only. 
Management authority and regular patrols of New Mexico segment of 
south fork of North McKittrick. Restricted access to Devil's Den. 

East Escarpment 

(A) Restricted access to Smith Canyon above Smith Spring. Boardwalk to 
protect Smith Spring from soil compaction and channel visitor traffic. 
Restricted access to Choza and Upper Pine springs. (B) Frequent patrol of 
Smith Spring to prevent foot traffic beyond fence. Regular patrol of Upper 
Pine and Choza springs to prevent camping. No camping areas near springs 
or smaller canyons. 



66 NORTHINGTON AND BURGESS 

Dog Canyon 

(A) and (B) Limit number and location of campers to aid recovery of 
meadows. No vehicular traffic beyond ranger station. 

Humphrey Canyon — PX Flat 

(A) and (B) Limited access. Day use only in vicinity of Sophora 
gypsophila. 

West Side 

(A) Route improved roads away from Jatropha dioica colony, gypsum 
dunes, and quartz sandhills. Boardwalk at Bone Spring to prevent trampling 
oi Astragalus and Lobelia. (B) No vehicular traffic off existing roads. Regu- 
lar patrol and maintenance of fences to prevent trespass grazing. Limited 
access to Bone Spring. 

SUMMARY 

The thrust of this report is the enumeration of all known plant taxa con- 
sidered unique within the park boundaries and management recommenda- 
tions designed to insure preservation of the communities. Although addi- 
tions, corrections, and revisions are anticipated with continuing field and 
laboratory work, the distribution, ecology, and significance of each of these 
taxa are as thorough and complete as presently possible. 

The management recommendations are based on our understanding of 
the plant communities in the park and are made in light of our observations 
of present visitor frequency and environmental variation. It must be 
reiterated that successful preservation of the park's unusual biota depends 
on continued monitoring and flexibility in management design. Unexpected 
environmental fluctuations and/ or vastly escalated public interest in the 
park could well render these recommendations inadequate, especially in cer- 
tain regions of the park. Continued survey and transect activity will result in 
our expected changes in the unique plant list found in this report. 



APPENDIX I 

Two codes are used in the appendix to permit more rapid assessment. Relative degree of 
uniqueness is given by Roman numerals as follows: 

I. Endemic to a relatively limited area in and/ or around the Guadalupe Mountains; 

II. Widespread in other states, but known in Texas from very few localities, usually 
known only from the Guadalupes; 

III. Of distributional interest; includes species with relatively small ranges and those near 
the limits of their known occurrence in the park; 

IV. Rare within the park, but may be common elsewhere in Texas; 

V. Not seen by the authors, but previously recorded from the park. 

The Arabic numerals give areas within the park in which the species has been found. The fol- 
lowing /ones have been defined primarily by watershed and elevation (Fig. 2). 

1. Lower McKittrick Canyon below Pratt Lodge; 

2. South McKittrick drainage below 7000 ft and including peak west of Turtle Rock; 

3. North McKittrick drainage including the New Mexico segment of the south fork of 
North McKittrick; 



STATUS OF PLANTS 



67 



4. South McKittrick drainage above 7000 tt excluding The Bowl; 

5. The Bowl drainage above the earth tank; 

6. Pine Spring Canyon drainage including Guadalupe Peak and El Capitan above 8000 it 
and the wash southwest of Uouser House; 

7. The Last Escarpment except McKittrick drainage; 

8. Upper Dog Canyon drainage; 

9. Humphrey Canyon drainage and PX Flat; 
10. Salt Basin drainage. 

Omissions from Appendix 1 

Several taxa are not included in this list because of a present lack of information. There are at 
least two species of Philadelphus in the park, including the endemic P. hitchcockianus\ how- 
ever, more study is needed to clarify distribution in the park. Four species of Symphoncarpos 
are recorded from the park. S. guadalupensis is endemic, but we have not yet obtained the 
flowering material critical for identification. Our understanding of sedges at this time is insuf- 
ficient to properly analyze our collections for Carex eburnea, a species found in the north- 
eastern United States, known in Texas only from the Guadalupe Mountains. Similarly, other 
notable species have not been reported here, but more information will be provided to the park 
service as it becomes available. 




Fig. 2. Distributional zones of plants in the Guadalupe Mountains National Park as 
referred to in Appendix I. 



68 NORTHINGTON AND BURGESS 

Name Category Zones 

Amelanchier utahensis Koehne II 2,3,4,6 

Scattered, high, forested slopes and riparian woodland. 
Andropogon hallii Hackel IV 10 

Appears restricted to gypsum dunes. Found immediately 

west of park boundary, probably rare in zone 10. 
Anthericum torreyi Baker III,(IV?) 4,8 

Scarce; open slopes above 6000 ft. 
Aquilegia chaplinei Standi. I 2,3 

Common, wet limestone crevices and gravel alluvium. 

Known only from Guadalupe Mountains. 
Asclepias tuberosa L. Ill 2,3,6,7 

Scattered in North and South McKittrick, Smith Canyon, 

and Pine Spring Canyon. Usually in sunlit pockets of 

soil on limestone ledges. Disjunct population of 

primarily eastern species. 
Aster hesperius Gray II 2,3,5 

Scattered among Cladium jamaicense along streams in 

McKittrick; scarce in meadows in The Bowl. 
Arbutus xalapensis H.B.K. Ill 1,2,3, 

Common in lower canyons throughout the park. 6,7,10 

A Mexican species near its northern limit. Does not 

appear to be reproducing well. 
Astragalus albulus Woot. & Standi. II 10 

Abundant at Bone Spring, not seen elsewhere. 

Only known occurrence in Texas. 
Astragalus pictiformis Barneby III 7,9 

Scattered on open slopes and meadows in Humphrey Canyon. 

A desert grassland species of southern New Mexico and 

northern Trans-Pecos Texas. 
Berberis repens Lindl. II 2,3,4, 

Widespread in forested areas. 5,6,8 

Bouteloua warnockii Gould & Kapadia III 7,9,10 

Common on dry limestone slopes. Considered rare over 

most of range (southern New Mexico to Coahuila). 
Caesalpinia jamesii (T. & G.) Fisher IV 10 

Uncommon; only on quartz sandhills. 
Campanula rotundifolia L. II 2,4,5 

Scattered on slopes and mesic limestone ledges. 
Castilleja latebracteata Penn. IV 2,3, 

Widespread but scarce at higher elevations. Generally 7,10 

on woodland slopes. 
Celastrus scandens L. II 2,3 

Small colony on slope and dry streambed of Devil's Den 

(0.5 km N, 1.1 km W Pratt Lodge); also a few plants near 

mouth of Devil's Den and south fork of North McKittrick, 

and South McKittrick. 
Centaur ea americana Nutt. IV 8 

Only seen on meadow north of Upper Dog Canyon Ranger 

Station. 



1,2,3, 


6,7,10 


2,6, 


7,10 



STAI US OF PLANTS 69 

Name Category Zones 

Cevallia sinuata Lag. IV 10 

Uncommon; limestone cliffs and dry washes in lower 

northwest corner of park. 
Chaetopappa hersheyi Blake 1 2,3,6,7 

Scattered populations in crevices of limestone cliffs. 

Usually above 6000 ft but extending lower in North and 

South McKittrick. 
Choisya dumosa (Torr.) Gray 111 

Common in lower canyons throughout park. Considered 

infrequent over most of Trans-Pecos (Correll and 

Johnston 1970). 
Chrysothammus nauseosus subsp. bigelovii (Gary) Hall & V 

Clem. 

Scattered on limestone ledges. 
Chrysothamnus pulchellus (Gray) Greene IV 7,10 

Previously collected in Smith Canyon (B. L. Turner 112, 

Sul Ross Herbarium). Uncommon on quartz, sandhills on 

west side. Analysis incomplete. 
Chrysothamnus spathulatus L. C. Anderson II 2,6, 

Open slopes. Analysis incomplete. 7,10 

Corallorhiza striata Lindl. Ill, IV 4 

Scarce; humus in forested canyon bottom approximately 

2.9 km N, 1.5 km E summit of Hunter Peak. 
Coreopsis lanceolata L. 11,111 2,3 

Identification tentative. A species of the eastern 

United States previously unreported from west Texas. 

Single specimen collected by B. H. Warnock (22804) in 

South McKittrick, labeled "infrequent." Collected in 

North McKittrick. 
Coryphantha dasyacantha (Engelm.) Orcutt. Ill 4,10 

Limestone ledges. Difficult to distinguish from more 

widespread C. strobiliformis. Specimens keying to this 

taxon collected on Blue Ridge and southwest of Cutoff 

Mountain. Probably occurs over most of west escarpment. 

Termed "endemic" in Correll and Johnston (1970). 
Coryphantha macromeris (Engelm.) Britt. & Rose IV 10 

Apparently uncommon. Open bajada and arroyos below 4500 ft. 
Coryphantha scheeri (O.Ktze.) L. Benson IV 10 

Very scarce. Gypseous soil near Lewis Well. 
Cystopteris hulbifera (L.) Bernh. 1 1, IV 2 

Previously collected at ". . . upper part of canyon near 

pools and rope swing" in 1964 (D. S. Correll 29803, 

SMU, UT Austin Herbaria). Material from The Narrows 

(approximately 2.6 km S, 2.0 km W Pratt Lodge) has been 

tentatively identified to this species. Locally common 

in shaded limestone crevices. 
Cystopteris fragilis (L.) Bernh. Ill 3,4,6 

Described as "local and rare" (Correll and Johnston 

1970). Uncommon, usually in shaded moist limestone 

crevices. 



70 NORTHINGTON AND BURGESS 



Name Category Zones 

Dalea frutescens Gray IV 1,2,7 

Apparently uncommon in park. Previously collected in 

Lower McKittrick (B. H. Warnock 9445, Sul Ross 

Herbarium). Locally occasional on slopes at Upper Pine 

Spring and in South McKittrick. 
Dalea seoparia Gray III 10 

Scattered on open areas of quartz and gypsum sand. 

More common on gypsum dunes immediately west of park. 

Probably at the eastern limits of its range here. 
Delphinium virescens Nutt. IV 9,10 

Previously collected a short distance east of the park, 

and to be expected in zone 7. A plains species probably 

near its western limit in Texas. 
Dicranocarpus parviflorus Gray III 10 

Restricted to gypseous soils. Can become abundant in 

years with good summer rains. Most common in Lewis 

Well Holmsley's Dugout area near west boundary. 
Epipactis gigantea Hook. IV 1,2,3 

Previously collected in South McKittrick (B. H. Warnock 

22811, 1968, Sul Ross Herbarium), labeled "infrequent." 

Seen by us at limestone seeps in New Mexico segment of 

south fork of North McKittrick, South McKittrick, 

and Lower McKittrick. 
Epithelantha micromeris (Engelm.) Weber III 2,3, 

Widespread in park, limestone ledges up to 7500 ft. 7,10 

Common in parts of North McKittrick. Reported to be 

heavily collected for sale in some parts of the state. 
Equisetum kansanum J. H. Schaffn. Ill 2,10 

Common along stream in parts of South McKittrick and 

at Bone Spring. 
Forestiera pubescens Nutt. Ill 2,4,8, 

Previously referred to as F. neomexicana\ Correll and 9,10 

Johnston (1970) consider Texas material to be a complex 

of intergrades between both taxa, and a more careful 

analysis is needed. Scattered and infrequent, often 

along streambeds. On west side a small colony occurs 

at Bone Spring. 
Fragaria bracteata Heller II 4,5 

Generally uncommon; forested slopes and canyons above 

7500 ft. 
Gaillardia multiceps Greene III 10 

Considered rare in Texas (Correll and Johnston 1970). 

Restricted to sandy gypsum soil on dunes and beach ridges 

along west boundary near Lewis Well. Occasional to 

uncommon. 
Glyceria striata (Lam.) Hitchc. II 2,3,7 

Uncommon. Scattered locations along streams in South 

McKittrick and Devil's Den. Small colony at Smith 

Spring. Rare in Texas. 
Grindelia havardii Steyerm. Ill 

Uncommon; open sites near streambeds. Type locality 

". . . dry gravelly wash near mouth of McKittrick Canyon" 



STATUS OF PLANTS 71 



Name Category Zones 

(J. A. Moore, J. A. Steyermark 283, 1931, UT Austin 

Herbarium). Limited distribution including Chisos 

Mountains; Crockett County, Texas; and New Mexico. 
Hackelia grisea (Woot. & Standi.) 1. M. Johnst. III,IV 3 

Apparently scarce; dry streambed in Devil's Den. 

Analysis of material incomplete, probably also in zones 

2 and 4. Occurs in mountains of Texas and New Mexico. 
Hedeoma apiculatum W. S. Stewart I 2,3,6 

Generally scarce, small colonies or isolated individuals 

in crevices of limestone cliffs. Bottom and slopes of 

South McKittrick, Devil's Den, south fork of North 

McKittrick, Hunter Peak, and higher parts of Pine 

Spring Canyon. Endemic to Guadalupe Mountains. 
Heterotheca viscida (Gray) Harms V 7 

Considered rare in Texas (Correll and Johnston 1970). 

Previously collected in Smith and Bear canyons (B. H. 

Warnock and M. C. Johnston 16583, Sul Ross Herbarium; 

B. L. Turner 1253, C. L. Lundell 14390, SMU). Analysis 

of our material incomplete. 
Hymenopappus biennis B. L. Turner V 2,4, 

Previously collected in South McKittrick, The Bowl, 5,6 

and Pine Spring Canyon (L. C. Hinckley 4472, B. H. 

Warnock 10997, M. C. Johnston 3168, Sul Ross; D. S. 

Correll 13920, C. H. Muller 8287 SMU; M. S. Young s.n., 

UT Austin). Analysis of our material incomplete. 

In Texas known only from Guadalupe Mountains. 
Hymenopappus tlavescens Gray IV 10 

Apparently rare in park. Collected just outside of west 

boundary on quartz sandhills. Previously collected near 

east boundary, and to be expected in zone 7. 
Hymenoxys richardsonii (Hook) Cockll. II 4,5,6 

Common in open areas throughout high parts of park. 

Southern limit of range. 
Ipomoea lindheimeri Gray III 1,3,7 

Widespread but scarce. Usually twining among low trees 

and shrubs. Endemic to west Texas. 
Ipomopsis arizonica (Greene) Wherry III, IV 4,5 

Scarce; meadows and open slopes above 6200 ft. May be 6,8 

more abundant after favorable rainfall distribution. 
Jatropha dioica Cerv. 1 1 1, IV 10 

Restricted to small colony on south slope of south 

Stagecoach Hill (0.8 km N, 3.5 km E Lewis Well). A 

species characteristic of the southern Chihuahuan Desert 

occurring here as the most northerly known disjunct. 
Juniperus scopulorum Sarg. (undescribed drooping form) (I?), Ill (7?), 8 

This species is scattered in mesic sites throughout higher 

parts of the park. Several individuals with a distinct 

drooping habit grow along streambed about 0.3 km S Upper 

Dog Ranger Station, and single individual observed at 

Upper Pine Spring. As a species, it is near southern 

limit of range here. 



72 NORTH1NGTON AND BURGESS 

Name Category Zones 

Lactuca graminifolia Michx. II 2,3 

Scarce, usually on open gravel alluvium. In summer 1974 

best growth observed in South McKittrick about 2.8 km S, 

1.7 km W Pratt Lodge. 
Lesquerella valida Greene 1 1,2, 

Apparently uncommon, but easily overlooked. Usually on 4,8 

open slopes. A poorly known species occurring in southern 

New Mexico, Guadalupe Mountains, and Sierra Diablo 

(Texas). 
Lilium philadelphicum L. II 2 ,3 

Only observed in mats of vegetation on limestone seeps, 

south fork of North McKittrick (New Mexico segment) and 

South McKittrick. Previously collected from South 

McKittrick (B. H. Warnock 22801, Sul Ross), labeled 

"rare." 
Lithospermum multiflorum Gray III 2,3,4, 

Widespread but often infrequent. Canyon bottoms; open 6,7 

slopes at higher elevations. In Texas known only from 

forested mountains in the Trans-Pecos. 
Lithospermum parksii I. M. Johnst. II-1II,IV 1,2, 

Generally scarce. Open or partially wooded slopes. May 6,7 

be confused with L. multiflorum. Species of limestone 

soils in West Texas. The variety rugulosum I. M. 

Johnst. is known from four localities; we have collected 

it near the top of Bear Canyon. 
Lithospermum viride Greene III 3,6,7 

Widespread and common. Usually in mesic, partially 

shaded sites. Probably also occurs in zones 2,4,5,8. 

Known from limestone soils in mountains of west Texas; 

ranges from New Mexico to Nuevo Leon. 
Lobelia cardinalis L. Ill 2,3, 

Common on sunny margins of permanent streams. On the 7,10 

east slope occurs at Smith, Choza, and Upper Pine springs. 

On west side seen only at Bone Spring. The varieties 

pseudosplendens McVaugh and multiflora (Paxt.) 

McVaugh appear to intergrade in the park, and in Texas 

both are known only from the Trans-Pecos. 
Lonicera arizonica Rehd. II 4,6 

Collected on forested limestone ledges about 0.5 km E 

summit of Guadalupe Peak and from South McKittrick. 

Analysis incomplete. Not known elsewhere in Texas. 
Machaeranthera blephariphylla (Gray) Shinners III 7,8,9 

Common, open or partially wooded limestone slopes 

between 5000 and 7500 ft. In Texas known only from 

limestone mountains in the Trans-Pecos. 
Mammillaria lasiacantha Engelm. Ill 10 

Apparently uncommon, but easily overlooked. Limestone 

ledges and alluvial fans below 6000 ft. A species occurring 

on limestone in the northeastern Chihuahuan Desert. 
Mentzelia humilis (Urban & Gilg.) Darl. Ill 10 

Scattered; restricted to gypsum soils and outcrops. 

Known only from gypsum in west Texas and New Mexico. 



STATUS OF PLANTS 73 

Name Category Zones 

Monolropa latisquama (Rydb.) Hult. Ill 4,5,7 

Generally uncommon, coniferous forest. In Texas known 

only from forested mountains of Trans-Pecos. 
Nama carnosum (Woot.) C. L. Hitchc. Ill 10 

Within park collected on gypsum outcrop 1.3 km S, 5.5 km 

W summit of Bush Mountain. Also seen immediately west 

of park boundary. A species restricted to gypsum soils 

of the northeastern Chihuahuan Desert. 
Nama xylopodum (Woot. & Standi.) C. L. Hitchc. 1 1,2,3,4, 

Widespread, limestone cliffs generally above 6000 ft. 5,6,7,10 

Endemic to limestone mountains of Trans-Pecos. 
Oenothera caespitosa Nutt. II 4,5 

Scarce, collected on limestone ledges on west side of 

Bush Mountain summit. Previously collected in similar 

habitat on Pine Top (B. H. Warnock and B. L. Turner 

181, Sul Ross). Probably at southern limit of its range 

here. 
Opuntia schottii Engelm. IV 10 

Found only in the lower northwest corner of the park, 

where it is locally common on alluvial fans. 
Oryzopsis hymenoides (R. & S.) Ricker III 10 

Infrequent, quartz and gypsum sand on the west side. 

Considered rare in Texas (Correll and Johnston 1970) 

which is eastern limit of range. 
Ostrya knowltonii Cov. II 2,3,4, 

Common in riparian woodland above 6000 ft, extending 6,7 

lower in McKittrick. Only known occurrence in Texas. 
Panicum ram ise turn Scribn. IV 10 

Uncommon on quartz sandhills. Probably near the western 

limit of its range here. 
Penstemon ambiguus Torr. IV 10 

Scarce on quartz sand. Characteristic of sandhills to the 

east of the park. 

Penstemon brevibarbatus Crosswhite III, IV 2,7,8 

Scarce; open grassy slopes. Collected in Pitchfork 
Canyon and north of Frijole. Previously found in South 
McKittrick ^B. H. Warnock and M. C. Johnston 16523, 
Sul Ross). Endemic to western Edwards Plateau and 
Trans-Pecos. 

Penstemon cardinalis Woot. & Standi, subsp. regalis I 2,3,4,7 

(A. Nels.) Nisbet & Jackson 

Generally uncommon; most abundant on limestone ledges 

and gravel alluvium in South McKittrick. Less frequent 

on limestone ledges in higher parts of Bear and Smith 

canyons. The subspecies regalis (A. Nels.) Nisbet & 

Jackson is considered endemic to the Guadalupe Mountains. 
Penstemon dasyphyllus Gray III 10 

Scattered on rocky slopes above 5500 ft W. of Guadalupe Peak. 

Previously thought to occur in Texas only in southern 

Trans-Pecos. 



74 NORTHINGTON AND BURGESS 

Name Category /ones 

Penstemon fendleri T. & G. IV 1,7 

Apparently scarce in park. Alluvial terraces in Lower 

McKittrick. Previously collected "1 mile west of 

Pine Spring along highway" (B. H. Warnock and M. C. 

Johnston 16302, Sul Ross). 
Perityle quinqueflora (Steyerm.) Shinners 1 2,3, 

Scattered populations on limestone cliffs; reaches 6,10 

greatest density in canyon bottoms between 5000 and 

7500 ft. Endemic to limestone areas of Trans-Pecos 

and adjacent New Mexico. 
Peteria scoparia Gray V(IV?) 7 

Previously collected in lower Pine Canyon (B. H. Warnock 

20837, Sul Ross), labeled "rare." Only two plants seen 

by us on the 9K Ranch about 20 mi. E of the park. Not 

seen in the park to date. 
Phanerophlebia auriculata Underw. 1 11, IV 2 

Only collected at The Narrows (about 2.6 km S, 2.0 km W 

Pratt Lodge) where it is uncommon in limestone crevices. 

Considered rare in the Trans-Pecos (Correll and 

Johnston 1970). 
Physocarpus monogynus (Torr.) Coult. II 2,4,5 

Generally scarce; seeps in South McKittrick and mesic 

slopes and canyons at higher elevations. Sole locality 

for Texas. 
Pinaropappus parvus Blake I 2,3,6, 

Scattered populations on limestone cliffs throughout 7,10 

higher elevations and extending lower along canyon 

bottoms. Endemic to northern Trans-Pecos and adjacent 

New Mexico. 
Polygala rimulicola Steyerm. 1 2,3,6, 

Scattered populations on limestone cliffs; generally 7,10 

above 6000 ft but occurs lower in McKittrick. Common 

in some areas. Endemic to Guadalupe Mountains and 

Sierra Diablo. 
Populus tremuloides Michx. 1 1, IV 4 

Small population at streambed junction about 2.5 km E of 

summit of Bush Mountain; consists of scattered trees in 

riparian forest of Pinus strobiformis, Pseudotsuga 

menziesii, Acer grandidentatum, and Ostrya knowltonii. 

Does not appear to be reproducing well, and probably will 

die out unless there is a local burn followed by protection 

from browsing. Another group reported for the park by C. 

Peterson (1973) about 4 km S, l'/ 2 km W Pratt Lodge. Not 

yet seen by us. 
Prosopis pubescens Benth. IV 10 

In the park known from three small trees on embankment 

of abandoned highway in Guadalupe Canyon, about 2.9 km S, 

2.5 km E summit of Guadalupe Peak. Probably introduced. 
Rhamnus smithii Greene (IV?), V 5 

Previously collected "near West bowl" (B. H. Warnock, 

22844, Sul Ross), labeled "infrequent." Not seen by us. 
Rhus toxicodendron L. IV 2,4 

Apparently scarce. Mesic sites in canyon bottoms. 



STATUS OF PLANTS 75 



Name C ategory Zones 

Robinia neomexicana Gray II 2,4,5,7 

Generally scarce; mesic, forested slopes and canyon 

bottoms, usually above 7000 ft. In Texas known only from 

Guadalupe and Franklin mountains (Correll and Johnston 

1970). 
Rosa stellata Woot. Ill 3,4,8, 

Widespread; often locally common. Open slopes above 9,10 

7000 ft, and on limestone ledges in North McKittrick and 

Devil's Den. Both the variety stellata and the variety 

earlansoniae W. H. Lewis have been collected within 

the park. More analysis is needed. 
Rosa woodsii Lindl. IILIV 2,4 

Scarce; collected among boulders near Turtle Rock and in 

forested canyon bottom northeast of The Bowl. 
Salvia farinacea Benth. Ill 1,2,3 

Occasional; mostly in open meadows and dry, gravelly 

streambeds. Probably near western limit of its range here. 
Salvia summa A. Nels. I 2,6, 

Generally scarce; more mesic sites on limestone cliffs; 7,10 

usually above 6500 ft. On west side found in a narrow 

canyon about 0.5 km N, 0.9 km W summit of Bush Mountain. 

Also collected in Bear and Smith canyons. 
Selinocarpus lanceolatus Woot. Ill 10 

Uncommon; restricted to gypsum soils. Limited distribution 

in Trans-Pecos Texas and New Mexico. 
Senecio neomexicanus Gray II 2,4,8 

Generally scarce; mejsic slopes and limestone ledges. Only 

known occurrence in Texas. May intergrade with another 

species in the park. Currently being studied by a specialist. 
Senecio warnockii Shinners I 10 

Scarce within park; restricted to gypsum. Only collected 

on gypsum outcrop about 1.2 km S, 5.5 km W summit of 

Bush Mountain; also seen near gypsum dunes just west of 

park boundary. Known only from Culberson and Hudspeth 

counties, Texas, and adjacent New Mexico. 
Sisyrhinchium demissum Greene II 2,7 

Uncommon; streamsides in South McKittrick. Previously 

collected at Smith Spring (B. H. Warnock 5405, Sul Ross). 

Sole occurrence in Texas. May be confused with S. 

ensigerum Bickn. which is more abundant in park. 
Smilacina racemosa (L.) Desf. Ill, IV 2,4,5 

Scarce; mesic forests. Previously collected in South 

McKittrick and The Bowl. Found by us in small canyon 

2.8 km N, 1.3 km E summit of Hunter Peak. In Texas 

known only from forested mountains of the Trans-Pecos. 
Solarium leptosepalum Correll III 4,8 

Uncommon, mesic canyon bottoms and slopes above 6500 ft. 

Probably near northern limit of range. 
Sophora gypsophila var. guadalupensis Turner & Powell I 9 

Common over a limited area south of Coyote Peak and in 

adjacent New Mexico. Otherwise known only from a small 

population on gypsum outcrop 20 mi. southwest of Coyame, 

Chihuahua, Mexico, and referred to there as var. 



76 NORTHINGTON AND BURGESS 

Name Category Zones 

gypsophila. Currently being studied intensively to 

determine relationship between populations. 
Sophora secundiflora (Ort.) DC. Ill 1,10 

Scattered over south-facing slope of Lower McKittrick, and 

in lower portions of canyons in the low-northwestern corner 

of park. Probably near the northern and western limit of 

range here. 
Stephanomeria wrightii Gray III 2,3,6 

Uncommon; usually on gravel alluvium in streambeds. 

Considered infrequent in the Trans-Pecos (Correll and 

Johnston 1970). 
Streptanthus sparsiflorus Roll. I 2,3,7 

Scarce; gravel alluvium and limestone ledges in canyon 

bottoms. In summer 1974, largest population seen in 

Devil's Den about 0.4 km N, 1.2 km W Pratt Lodge. Very 

scarce in upper Smith Canyon. Previously collected near 

summit of Pine Top (G. L. Webster 4595, Sul Ross). 

Endemic to Guadalupe Mountains. Two other species (S. 

platycarpus Gray and S. carinatus Wright) have been 

reported from park, and may be confused. Abundance 

highly variable; dependent on seasonal rainfall. 
Swert ia radiata (Kell.) O. Ktze. II 3,4, 

Scattered throughout high elevations; best growth observed 5,7 

in mesic canyon bottoms. In Texas known only from 

Guadalupe Mountains. 
Valeriana arizonica Gray 114 

Scattered populations on limestone outcrops in mesic 

canyons. 
Valeriana texana Steyerm. I 2,3,4 

Widespread but generally uncommon; more mesic crevices 6,7 

in limestone cliffs and ledges above 6000 ft. Endemic to 

Guadalupe Mountains. 
Viguiera multiflora (Nutt.) Blake II 4 

Single specimen collected on ridgetop about 2.3 km S, 

0.6 km E summit of Lost Peak. Identification tentative, 

but if confirmed, represents first record from Texas. 
Viola missouriensis Greene II 2,3,7 

Uncommon; greatest density observed on limestone ledges 

of South McKittrick below The Narrows. Scarce in crevices 

of boulders at Smith Spring. Occurs here as a marginal 

disjunct population. 
Yucca faxoniana Sarg. Ill, IV 6,7,10 

Scarce; limestone ledges and alluvial fans between 5000 

and 6700 ft. Greatest density observed on ridge south of 

Smith Spring. Small colonies on north side of mouth of 

Pine Spring Canyon and south rim of Bone Canyon. Probably 

near northern limit of range here. Large individuals in 

Lower McKittrick referred to Y. carnerosana (Trel.) 

McKelvey in Correll and Johnston 1970. 
Zigadenus elegans Pursh. II 2,3 

Uncommon; in mats of vegetation at seeps. Sole known 

occurrence in Texas. 



STATUS OF PLANTS 77 

REFERENCES 

Pertinent literature relative to unique plant taxa from the Guadalupe Mountains 
region are given below. 

Correll, D. S., and M. C. Johnston. 1970. Manual of the Vascular Plants of 

Texas. Tex. Res. Found., Renner, Texas, 1881 pp. 
Gehlbach, F. R.,B. H. WarnocicW. C. Martin, and H. K.Sharsmith. 1969. 

Vascular plants of Carlsbad Caverns National Park, New Mexico and adjacent 

Guadalupe Mountains (N.M. -Texas). Revised checklist compiled by Carlsbad 

Caverns National Park, Carlsbad, New Mexico. 
Gould, F. W., ed. 1973. Preliminary TOES list of Texas rare, endangered and 

marginal plant species. Checklist from Texas Organization of Endangered 

Species, Texas A&M Univ. 
Johnston, M. C. 1971 (Fall) 1972 (Spring). Rare and endangered plants native to 

Texas. Rare Plant Study Center, P.O. Box 8495, Austin, Texas 78712. 
Riskind, D. H. 1973. The rare and endangered flora of the Guadalupe Mountains 

National Park. Report submitted to the Chief Biologist, Southwest Region Na- 
tional Park Service, Santa Fe, New Mexico 87501. 
Turner, B. L., and A. M. Powell. 1972. A new gypsophilic Sophora 

(Leguminosae) from northcentral Mexico and adjacent Texas. Phytologia 

22:419-423. 
Vines, R. A. 1960. Trees, Shrubs and Woody Vines of the Southwest. Univ. Texas 

Press, Austin, 1104 pp. 
Weniger, D. 1969. Cacti of the Southwest. Univ. Texas Press, Austin, 249 pp. 



ACKNOWLEDGMENTS 

We would like to acknowledge gratefully the cooperation of the Guada- 
lupe Mountains National Park staff during this study. Acknowledgment is 
also made to L. T. Green and Tim Holland for their field assistance. Field 
work was supported by the Department of the Interior, National Park 
Service, grant number CX-700040145. 



Agave — Complex of the Guadalupe 
Mountains National Park; Putative 
Hybridization Between Members of 
Different Subgenera 



TONY L. BURGESS, Texas Tech University, Lubbock 

When I began surveying the flora of Guadalupe Mountains National 
Park, my observations of the diversity of Agave in McKittrick Canyon 
prompted me to investigate the variation within this group. Two distinct 
taxa appeared to be present along with a third group which appeared 
morphologically intermediate. I naturally formed the hypothesis that the 
third group was of hybrid origin. 

Field collections began in early summer of 1973 and are continuing 
through the present. Due to the paucity of material found in 1974 caused by 
drought early in the year, I have limited the analysis presented here to the 
first year's collection which is comprised of flowers and leaves of 46 plants 
from Guadalupe Mountains National Park, Texas. 

TAXONOMY 

Specimens collected during the first year were identified with the key pre- 
pared by Dr. Howard Scott Gentry for the Manual of the Vascular Plants of 
Texas (Correll and Johnston 1970). According to the key, five species were 
collected within the park — Agave lecheguilla, A. chisosensis, A. gracilipes, 
A. neomexicana, and A. parryi. 

Agave lecheguilla, in the subgenus Littaea, is easily recognized by its 
smaller size and spike-like inflorescence (Fig. 1). It is often a dominant on 
xeric rocky slopes below 5500 ft elevation throughout the park. As a species, 
it occurs from arid north-central portions of Mexico north through New 
Mexico, and is often considered a characteristic plant of the Chihuahuan 
Desert. Particularly robust specimens, including a form with yellow flowers, 
can be found in Lower McKittrick Canyon. 

Populations of a larger Agave in the subgenus Euagave, section 
Parrayanae occur primarily on open sites from about 5500 ft elevation to the 
highest parts of the range above 8000 ft. Plants which key to A. parryi on the 

79 



80 



BURGESS 



±4 



>wS 



/If 







♦ 



Fig. 1. /Igave lecheguilla in Lower McKittrick Canyon, Guadalupe 
Mountains. 



basis of their relatively larger leaves were found in open woodland on the 
slopes of McKittrick Canyon; however, most plants collected from higher 
parts of the park were identified as A. neomexicana. A. parry i was de- 
scribed by Englemann ( 1 875) who considered it to range from northern Ari- 
zona through western New Mexico, ". . . and perhaps eastward to the 
mountains below El Paso." A.neomexicana was described from the Organ 
Mountains in south-central New Mexico by Wooten and Standley (1913). 
Dr. Gentry has provided much needed help in the classification of this 
complex. After his trip to the Guadalupes in 1974, Dr. Gentry kindly wrote 
to explain that, based on his observations, A. parryi does not occur east of 



AGAVE-COMPLEX 81 



%> 



*S* 







Fig. 2. Agave neomexicana in Lower McKittrick Canyon, Guadalupe 
Mountains. 



the Rio Grande. However, he has pointed out that the type of A. havardiana, 
another species in the section Parrayanae, was cited by Trelease (191 1) as 
"Guadaloupe Mountains" from a collection by Harvard in 1 88 1 . Although I 
have seen no plants large enough to fit with this species inside the park, Dr. 
Gentry has found a single A. havardiana on the bajada east of the Guada- 
lupe Mountains, and its presence in the park is possible. Based on these con- 
siderations, the larger form of Agave in the Guadalupe Mountains (Fig. 2) 
will be called A. neomexicana throughout this report. Caution is advisable in 
applying this name indiscriminately, for A. gracilipes is also present in the 
park, and is indistinguishable from A. neomexicana on the basis of leaves 



82 



BURGESS 



alone. Dr. Gentry has advised me that the major differences between the two 
taxa are relative length of the floral tube (1/3 as long as tepais in A. 
gracilipes, 1 / 2 to 2/ 3 as long in A. neomexicana), and season of bloom (early 
summer for A. neomexicana, autumn for A. gracilipes). In 1973, the 
McKittrick Canyon plants bloomed in June and July, whereas in higher 
parts of the range flowering continued through August. Most later- 
blooming plants could still be included with A. neomexicana on the basis of 
floral tube length; however, several plants collected in late August from the 
area north of Lost Peak had floral tubes short enough to be called A. 
gracilipes, even though with respect to all other characters examined they 
were closer to A. neomexicana. 



W 

X - 



1 

















Fig. 3. Putative hybrid Agave (A. gracilipes) in Lower McKittrick Canyon, 
Guadalupe Mountains. 



AGAVE-COMPLEX 83 

Some confusion may exist in delimiting A. gracilipes and A. chisosensis. 
The former species was described from Sierra Blanca in Hudspeth County, 
Texas (Trelease 1911), and the latter was first collected in the Chisos Moun- 
tains of Brewster County, Texas (Muller 1939). Gentry, in Correll and 
Johnston ( 1970), differentiates between the two taxa on the basis of panicle 
structure. In A. gracilipes it is **. . . usually narrow and closely branched with 
20 to 30 lateral branches in upper half to third of shaft . . . ," whereas in A. 
chisosensis the panicle is described as ". . . racemose, 5-6 m. tall; flowers 
short-pedicillate, borne in compact clusters of 10 to 12 . . . ." C. H. Muller's 
original description (1939) of A. chisosensis reads as follows: "Panicle very 
strict and spike-like, its branches 4-5 cm. long, densely flowered. . . ."Floral 
characters as described for both species are generally similar; however, there 
are differences in dimensions. A. chisosensis is larger with respect to all 
dimensions given except the length of the ovary. The only other difference of 
note appears to be the ratio of tepal width to tepal length, being 1:3 to 1:5 in 
A. gracilipes, and about 1:3 in A. chisosensis. There appear to be two main 
differences in leaf morphology between the species as described. In A. 
chisosensis the leaves are over 50 cm long, and the corneous margin is con- 
tinuous from the base to the terminal spine. Leaves of A. gracilipes are under 



TABLE 1. Mean and standard deviat 


ion of variables 


measured for each 


Agave taxon. 


Variable 


A. lecheguilla 


A. neomexicanc 


i A. "gracilipes" 


Inflorescence 








Total height (dm) 


28.2, 7.5 


34.9, 5.8 


37.4, 5.7 


Mid-panicle branch 








length (cm) 


3.0, 1.5 


27.8, 9.4 


10.6, 8.1 


Number of panicle branches 


72, 24.2 


14, 4.2 


39, 22.5 


Leaf 








Length (cm) 


28.4, 6.8 


26.5, 6.2 


32.1,6.4 


Maximum width (cm) 


3.2, 0.7 


8.6, 1.2 


6.2, 1.3 


Maximum marginal tooth (mm) 


5.4, 1.5 


5.0, 1.1 


5.3, 2.0 


Marginal tooth at widest 








part of blade (mm) 


3.7, 1.7 


1.7, 1.1 


2.9, 1.9 


Corneous margin 








length (mm) 


254, 66 


72, 22 


164, 95 


Terminal spine length (mm) 


25.9, 3.2 


34.2, 9.4 


29.4, 9.2 


Flowers (all in mm) 








Ovary body 


14.2, 2.1 


26.2, 5.3 


19.1, 3.6 


Ovary neck 


4.6, 1.1 


5.2, 1.5 


5.6, 2.0 


Tube length 


3.5, 1.0 


12.7, 2.4 


7.0, 1.8 


Filament insertion 


3.2, 1.1 


8.9, 2.0 


5.1, 1.1 


Maximum tepal length 


14.9, 2.0 


15.8, 1.9 


16.1,2.0 


Minimum tepal length 


13.6, 2.0 


13.7, 1.7 


14.3, 1.9 


Outer tepal width 


4.0, 0.7 


5.7, 1.0 


4.9, 0.7 


Filament length 


34.7, 6.0 


42.1,7.1 


38.5, 6.7 


Anther length 


14.1, 1.9 


19.6, 2.5 


16.9, 2.1 



84 BURGESS 

30 cm in length, and the terminal spine is decurrent at most to about the mid- 
dle of the blade. 

The putative hybrid form (Fig. 3) appears close to the above two species. 
Plants were found blooming from July through October, and several 
panicles were killed by frost prior to capsule maturation. The greatest con- 
centration of this taxon was observed in McKittrick Canyon on alluvial ter- 
races and open slopes mostly between 5000 and 6000 ft elevation. Other 
plants which could be included in this group were seen on the bajada east of 
the park, on a small hill southwest of El Capitan, and at Juniper Well in Lin- 
coln National Forest north of the park. Mean leaf length of the putative 
hybrid specimens was 32 cm, closer to A. gracilipes; furthermore, the mean 
length of the corneous margin from the terminal spine was half the mean 
total leaf length, another A. gracilipes character. All mean floral dimen- 
sions were close to those described for A. gracilipes, but the tepal 
width: length ratio was an inconclusive 1:3. Panicle structure was variable as 
to number and length of branches (Table 1), but the mean branch length was 
longer than that described for A. chisosensis. Based on this analysis, I have 
concluded that the putative hybrid can probably be grouped with A. gra- 
cilipes if a name must be given. 

METHODS AND RESULTS 

Several approaches are desirable for investigating patterns of variation 
because each different avenue increases the perspective. The most obvious, 
and probably most important, is the analysis of morphology. Other possi- 
bilities for clarifying relationships include data on phytochemistry, 
karyotypes, and ecological parameters. At this time only morphology has 
been studied enough to warrant presentation, although data are being 
gathered on all the above aspects for later summarization. 

Measurements of the following characters were taken from two leaves per 
plant: total length, maximum width, number of marginal teeth on each side, 
maximum marginal tooth length, length of marginal teeth on each side of the 
widest part of the leaf, length of corneous margin decurrent from the base of 
the terminal spine, and length of terminal spine. The floral measurements 
used by Dr. Gentry ( 1 972) were adopted for this study. Ten flowers per plant, 
when available, were evaulated with respect to the following variables: 
length of ovary body, length of ovary neck, length of floral tube, length from 
bottom of the floral tube to point of filament insertion, maximum tepal 
length, minimum tepal length, width of outer tepal, filament length, and 
anther length. 

Leaf and flower data were analyzed separately by computer. Overall sta- 
tistics, analysis of variance, and a principal components analysis were done 
using the program SIMPLE developed by Dr. Charles Gaskins at Texas 
Tech University. A discriminant analysis was done with the program BMD 
07M, part of a package of biomedical computer programs from the Univer- 
sity of California (Dixon 1971). Statistical analysis by computer generally 



,4(7/4 ^-COMPLEX 



85 



confirmed the conclusion intuitively reached in the field; that the putative 
hybrid form bridged a morphological gap between A. neomexicana and A. 
lecheguilla. Principal components analysis of overall variation showed no 
clear trends in the relative contributions of the variables. An analysis of vari- 
ance indicated that the variation of floral dimensions within a plant was 
much less than between plants, as expected. Two strategies were used in the 
BMD 07M stepwise discriminant analysis. In one strategy a classification 
scheme was derived to distinguish between A. lecheguilla and A. neomexi- 
cana specimens, then data from the putative hybrid were classified by the 
same scheme. The other strategy considered each group a separate entity, 
and a scheme was derived to discriminate among all three groups. 

Discriminant analysis of leaf data showed that width, total length, and 
length of corneous margin were the best characters to distinguish among the 
three groups. Group means for all variables are presented in Table 1. When 
treated as a separate group, the putative hybrid was in an intermediate posi- 
tion morphologically and appeared to be contiguous or overlap slightly with 
A. neomexicana and A. lecheguilla. In the two-group strategy, the putative 
hybrid was somewhat closer to A. lecheguilla. 




Fig. 4. Scanning electron micrograph of cuticle surface from Agave neomexicana, 
500 x. 



86 



BURGESS 




Fig. 5. Scanning electron micrograph of cuticle surface from Agave lecheguilla, 
500 x. 



Based on data from floral characters, discriminant analysis using the 
three-group strategy again showed the putative hybrid in an intermediate 
position, somewhat closer to A. lecheguilla but merging into the other two 
species. Results from the two-group strategy had the same pattern. 
Characters which could best distinguish among the three groups were, in 
order of decreasing importance, floral tube length, length from bottom of 
tube to filament insertion, and width of outer tepal. 

Flavonoid analysis of leaf extracts was done by two-dimensional paper 
chromatography as described by Mabry et al. (1970). Initial results were 
promising, but the pattern of variation rapidly became confusing. Experi- 
mentation has indicated differences in flavonoid patterns between flower- 
ing and immature rosettes of the same clone. In addition, changes in com- 
pound patterns resulting from slow drying of intact leaves seem probable. 
Investigation of seasonal variation in flavonoid pattern is currently under- 
way. Thus much work remains before an analysis of flavonoids in Agave can 
be interpreted properly. 

Blunden and Jewers (1973) have proposed that cuticle and stomate 
anatomy could be used to differentiate species of Agave. Comparison of 
photographs taken by scanning electron microscope (Figs. 4-6) shows dis- 



AGAVE COMPLEX 87 










Fig. 6. Scanning electron micrograph of cuticle surface from putative hybrid 
Agave shown in Fig. 3, 500 *. 



tinct differences in surface texture and shape of cuticular cells from A. neo- 
mexicana and A. lecheguilla. Notice also that the cuticle of a putative hybrid 
appears intermediate between the two species. Conclusions should not be 
made from these single specimens, but they indicate a promising source of 
information for Agave systematics. 

DISCUSSION 

Calling the intermediate form Agave gracilipes does not negate the possi- 
bility of hybrid status. Morphological patterns for the three groups indicate 
that locally A. gracilipes occupies a position morphologically if not 
genetically between A. lecheguilla and A. neomexicana. Furthermore, the 
lack of sharp morphological discontinuity between the three taxa probably 
reflects a similar genetic situation, so that even if A. gracilipes is not itself a 
hybrid, there are hybrids between that taxon and the other two species. 
Studies of karyotypes in the genus show a relatively stable base number of n 
- 30 and indicate a strong possibility of alloploidy and autoploidy (Granick 
1944). The only chromosome counts reported for species occurring in the 
Guadalupe Mountains are In - 55, 60, and 120 for A. lecheguilla (Granick 
1944; Cave 1964). The only apparent barrier to cross pollination is time of 



88 BURGESS 

bloom, and there is considerable overlap among the three taxa. The large 
amount of nectar attracts a variety of hummingbirds, wasps, bees, butter- 
flies, and beetles; and it seems reasonable to postulate a nonspecific pollina- 
tion strategy. The late bloom of the putative hybrid is perplexing because it 
gives these plants a decided disadvantage in sexual reproduction under the 
current regional climate. This could help account for the relatively limited 
distribution of this form in the region. This characteristic is clearly different 
from both proposed parental species, and could signify genes not currently 
present in other local taxa. However, until the mechanism of bloom induc- 
tion is known, it is difficult to evaluate the genetic significance of this factor. 
In any case, asexual cloning appears to be an effective propagating strategy 
on a local basis. 

To conclude, morphological data from 1973 collections of Agave in 
Guadalupe Mountains National Park show a good possibility of hybrid 
origin for some populations tentatively included with A. gracilipes. There 
appears to be backcrossing with the two proposed parental species — A. 
neomexicana and A. lecheguilla. However, I can make no firm conclusions 
without more data, which I am continuing to gather for future analysis. 

LITERATURE CITED 

Blunden, G., and K. J ewers. 1973. The comparative leaf anatom> of Agave, 

Beschomeria, Doryanthes, and Furcraea species (Agavaceae: Agaveae). Bot. J. 

Linn. Soc. 66:157-179. 
Cave, M. S. 1964. Cytological observations on some genera of the Agavaceae. 

Madrono 17:163-170. 
Correll, D. S. and M. C. Johnston. 1970. Manual of the Vascular Plants of 

Texas, Tex. Res. Found. Renner, 1881 pp. 
Dixon, W. J., ed. 1971. BMD Biomedical Computer Programs. Univ. California 

Press, Berkeley, 2nd ed. revised, 600 pp. 
Engelmann, G. 1875. Notes on Agave. Trans. St. Louis Acad. Sci. 3:291-322. 
Gentry, H. S. 1972. The Agave family in Sonora. Agric. Handbook No. 399, 

Agric. Res. Serv. U.S. Dept. Agric, U.S. Government Printing Office. D.C., 195 

pp. 
Granick, E. B. 1944. A karyosystematic study of the genus Agave. Am. J. Bot. 

31:283-298. 
Mabry,T. J. MARKHAM.and M. B. Thomas. 1970. The Systematic Identification 

of Flavonoids. Springer-Verlag, New York, 354 pp. 
Muller, C. H. 1939. A new species of Agave from Trans- Pecos Texas. Am. Midi. 

Nat. 21:763-765. 
Trelease, W. 1911. Revision of the Agaves of the group Applanatae. Rep. Mo. 

Bot. Gard. 22:85-97. 
Wooten, E. O., and E. P. Standley. 1913. Descriptions of new plants 

preliminary to a report upon the flora of New Mexico. Contrib. U.S. Natl. Herb. 

16:109-196. 



AGAVE-COMPLEX 



89 



ACKNOWLEDGMENTS 

I am particularly indebted to Drs. Howard Scott Gentry and David K. 
Northington for encouragement and advice. The staff of Guadalupe Moun- 
tains National Park and Fred E. Lyons were most helpful in the initiation of 
field work, which was supported by NPS grant number CX-700040145 and 
The Museum of Texas Tech. Equipment and materials for flavonoid 
analysis were supplied by the Institute of Environmental Chemistry. 



The Land and Freshwater Mollusca 
of the Guadalupe Mountains 
National Park, Texas 



RICHARD W. FULLINGTON, Dallas Museum of Natural 
History, Texas 

On 1 November 1922, H. A. Pilsbry (often considered the father of south- 
western malacology) and James H. Ferriss hiked from the west along the old 
PX trail into the Texas Guadalupe Mountains, thus making the first 
recorded attempt to collect mollusks in the Permian Reef. They had started 
their hike in Orange, New Mexico (which no longer exists), and during their 
entire 1 1-day trip, Pilsbry believed that they were collecting in New Mexico. 
Their search carried them to areas across the west side, up to the area of Bush 
Mountain and up Pine Springs Canyon twice. Pilsbry (1939, 1946) later 
recorded only the larger species collected and none of the smaller species that 
they surely gathered. Perhaps more strangely, though, is that after this 
beginning, no other major molluscan collecting efforts were made in the 
Guadalupe Mountains until 1974. 

Pilsbry did return for one day on 18 July 1935 with Cyril Harvey; they 
climbed to about 6500 ft on the southern flank of "Signal" (Guadalupe) 
Peak. Metcalf (1970) published Pilsbry's field notes of these trips. In 1934 
and 1935, H. C. Fountain, who aided P. B. King during his geological sur- 
vey of the Guadalupes of Texas, collected four species of living snails from 
Pine Springs Canyon and a series of fossil snails from alluvial deposits in Bell 
Canyon and in Pine Springs Canyon. The material was identified by Dr. 
J. P. E. Morrison of the U.S. National Museum, and the list was published 
in King's (1948) Geology of the southern Guadalupe Mountains. 

In July 1950, Dr. E. P. Cheatum with his wife Edith and their two small 
sons visited the Texas Guadalupes to collect Mollusca. Shortly after arriving 
at the small motel at Pine Springs, Texas, Dr. Cheatum became ill and was 
confined to the room for most of their stay which, historically, was unfortu- 
nate as will be explained in the discussion of Ashmunella edithae. He did 
make one excursion into McKittrick Canyon but the treks up the moun- 
tains were made by his wife and eldest son who were accompanied by a group 

91 



92 FULLINGTON 

of geologists. The results of this trip were published by Pilsbry and Cheatum 
(1951). 

Mr. Munroe L. Walton of Glendale, California, collected in Pine Springs 
Canyon on 19 March 1952 and 23 October 1952, and published (1963) on 
one of the two collected species and sent the other to Dr. J. C. Bequaert. 

In April 1958 and April 1961, Dr. Joseph C. Bequaert collected a series of 
small mollusks from washed up debris in Bell Canyon at the Highway 180 
crossing and live Physa in the bed of the canyon. He visited McKittrick Can- 
yon on 4 and 5 April 1966, where he collected another series of small gastro- 
pods. Soon after, he presented the National Park Service an unpublished 
checklist of all known mollusks collected in the Park. The manuscript 
(Bequaert 1966) was made available to me for incorporation into this report. 

Other investigators periodically have collected mollusks in the Park, 
including Dr. Fred Gehlbach, Baylor University, who informed Dr. 
Bequaert that over a 10-year period while doing ecological work in the Park, 
he had collected gastropods. These were deposited in the University of 
Michigan, Museum of Zoology, but no publications have resulted. 

Dr. Artie L. Metcalf, University of Texas at El Paso, has worked 
periodically on the Guadalupe Mollusca. He graciously has made much of 
his data available to me and has a manuscript in preparation listing species 
he collected in Pine Springs Canyon, with a description of a new fossil 
Ashmunella. 

Mr. Lloyd E. Logan, Texas Tech University, while investigating Quater- 
nary vertebrate remains from Sloth and Dust caves, sifted fossil gastropods 
from floor sediments. The material was sent to me for identification, and Mr. 
Logan kindly permitted me to incorporate the species list into this report. 

My first trip to the Texas Guadalupes was made with Dr. E. P. Cheatum 
and Mr. Hal P. Kirby in November 1969. We collected at Smith Spring and 
McKittrick Canyon. In November 1970, the three of us returned accom- 
panied by William E. Wilson. Wilson and I worked up Pine Springs Can- 
yon, Pipeline Canyon, The Bowl, and up the south walls of McKittrick Can- 
yon where we collected Ashmunella carlsbadensis for the first time in Texas. 
Cheatum and Kirby worked the lower levels. Dr. Cheatum, Wilson, and I 
made a brief stop in McKittrick Canyon in November 1971. All of the mate- 
rial collected thus far was published with the molluscan data that Mr. Lloyd 
Pratt, Fort Worth Museum of Science and History, had gathered from the 
Chisos Mountains and west Texas in general (Cheatum et al. 1972). 

Fourteen members of the museum staff spent 2 weeks in July 1974 col- 
lecting mainly at the upper elevations. Main camp was established at the 
ranger's headquarters in Upper Dog Canyon and we backpacked into the 
mountains from there. The "rainy season" had begun which greatly aided 
collecting as most of the gastropods were above ground. I returned in 
December 1974 with staff members Walt Davis, Bill Wilson, and Steven 
RunneL, at which time we collected in Pine Springs Canyon, Bone Canyon, 
along the summit of the eastern escarpment, and near most of the springs 



MOLLUSCA 93 

along the east side of the mountains. Also, during this last trip, we collected 
to the head of Pine Springs Canyon. 

The bulk of this report is the data gathered from these two intensive trips 
in 1974. The collecting sites are listed below, grouped by altitude intervals. 
Lesser known sites are indicated by degrees, minutes, and seconds from the 
U.S. Geological Survey Guadalupe Peak, Texas, Quadrangle map, 1933 
edition. 

COLLECTING SITES 1974 

5000 to 5500 ft (1500 to 1650 m) 

floor of McKittrick and South McKittrick canyons 

Choza Spring 

Manzanita Spring 

Juniper Spring 

Bone Canyon 
5500 to 6000 ft (1650 to 1800 m) 

Lower Pine Springs Canyon (to first big bend) 

Smith Spring 

Spring at base of Pipeline Canyon 
6000 to 6500 ft (1800 to 1950 m) 

104° 50' W, 31° 59' 30" S (ravine just west of Ranger Station in Upper Dog Canyon) 

104° 51' W, 31° 59' 30" S (trail to Coyote Peak) 

104° 50' E, 31° 59' N (canyon heading on east ridge of Upper Dog Canyon by Devil's Den 
Canyon) 

104° 50' W, 31° 59' 30" S 

104° 50' 30" E, 31° 58' 30" N 

Pine Springs Canyon 
6500 to 7000 ft (1950 to 2100 m) 

104° 50' 30" E, 31° 58' S ("Spring" Canyon off Lost Peak Trail) 

104° 50' 20" W, 31° 57' 17" N 

Upper Portion of Pine Springs Canyon 

Lower Pipeline Canyon 
7000 to 7500 ft (2100 to 2500 m) 

104° 49' 31" W, 31° 59' N 

104° 50' E, 31° 57'30"N 

104° 51' W, 31° 56' 30" N 

104° 50' 35" W, 31° 56' 20" N 

104° 51' W, 31° 55' 30" N 

Pine Springs Canyon 
7500 to 8000 ft (2250 to 2400 m) 

Pipeline Canyon 

Pine Springs Canyon 

104° 51' 20" W, 31° 55' 45" N (Aspen Grove) 

104° 51' 30" W, 31° 56'35"N 

104° 50' 20" W, 31° 57' 17" N 

Lost Peak 
8000 to 8300 ft (2400 to 2490 m) 

104° 52' W, 31° 56' 40" N (Blue Ridge Trail area) 

104° 52' 20" W, 31° 56' 55" N (Blue Ridge Trail area) 

104° 52' 40" W, 31° 56' 55" N (Blue Ridge Trail area) 

104° 51' 35" W, 31° 56' 35" S 

East escarpment to just beyond Smith Spring Canyon 

The Bowl (southern area only) 



94 FULL1NGTON 

Checklist of Land and Freshwater Mollusca of the 
Guadalupe Mountains National Park, Texas 

This list represents all published and unpublished species for the Texas Guadalupe Moun- 
tains. Strictly indigenous species are noted as are species that are known currently only as 
fossils. All current taxonomic changes are also incorporated and are explained in the sys- 
tematic discussion section. Numbers in parentheses indicate species included in each group. 

Pelecypoda (1) 
Family Sphaeriidae 

Pisidium casertanum (Poli) 
Gastropoda (aquatic) (5) 
Family Hydrobiidae 

Hydrobia sp. 
Family Lymnaeidae 

Lymnaea humilus (Say) 
Family Physidae 

Physa virgata (Gould) 
Family Planorbidae 

Helisoma trivolvis trivolvis (Say) 

Gyraulus parvus (Say) 
Gastropoda (terrestrial) (55) 
Family Helminthoglyptidae 

Humboldtiana ultima Pilsbry 
Family Urocoptidae 

Metastoma roemeri roemeri (Pfeiffer) 

Holospira montivaga montivaga Pilsbry (indigenous) 

Holospira montivaga form brevaria Pilsbry (indigenous) 

Holospira pityis Pilsbry and Cheatum (indigenous) 

Holospira oritis Pilsbry and Cheatum (indigenous) 

Holospira danielsi Pilsbry and Ferriss 
Family Bulimulidae 

Rabdotus dealbatus dealbatus (Say) 

Rabdotus dealbatus durangoanus (Martens) 
Family Thysanophoridae (Sagdidae) 

Thysanophora hornii (Gabb) 
Family Endodontidae 

Helicodiscus nummus (Vanatta) 

Helicodiscus singleyanus singleyanus (Pilsbry) 

Helicodiscus eigenmanni Pilsbry 

Discus cronkhitei (Newcomb) 

Punctum minutissimum (Lea) 

Punctum vitreum H. B. Baker 
Family Zonitidae 

Zonitoides arboreus (Say) 

Glyphyalinia indentata paucilirata (Morelet) 

Nesovitrea electrina (Gould) 

Nesovitrea sp. 

Euconulus fulvus ( M iille r) 

Striatura meridionalis (Pilsbry and Ferriss) 

Vitrina pellucida alaskana Dall 

Hawaiia minuscula minuscula (Binney) 

Hawaiia minuscula neomexicana (Cockerell and Pilsbry) 



MOLLUSCA 95 



Family Limacidae 

Deroceras laeve (Muller) 
Family Polygyridae 

Ashmunella kochi amblya Pilsbry 

Ashmunella rhyssa rhyssa (Dall) (fossil only) 

Ashmunella carlsbadensis Pilsbry 

Ashmunella edithae Pilsbry and Cheatum (indigenous) 

Ashmunella sp. nov. 
Family Succineidae 

Succinea luteola Gould 

Succinea sp. 
Family Cionellidae 

Cionella lubricella (Porro) 
Family Pupillidae 

Gastrocopta armifera (Say) 

Gastrocopta procera procera (Gould) 

Gastrocopta pentodon (Say) 

Gastrocopta pilsbryana (Sterki) 

Gastrocopta ashmuni (Sterki) 

Gastrocopta contracta (Say) 

Gastrocopta pellucida (Pfeiffer) 

Pupilla muscorum muscorum (Linne) (fossil only) 

Pupilla blandi Morse 

Pupilla sonorana (Sterki) 

Pupoides albilabris (C. B. Adams) 

Pupoides hordaceus (Gabb) (fossil only) 

Vertigo milium (Gould) 

Vertigo gouldii arizonensis Pilsbry & Vanatta 

Vertigo ovata (Say) 
Family Valloniidae 

Vallonia perspectiva Sterki 

Vallonia gracilicosta Reinhardt 

Vallonia parvula Sterki (fossil only) 

Vallonia cyclophorella Sterki (fossil only) 
Family Strobilopsidae 

Strobilops labyrinthica (Say) 
Family Oreohelicidae 

Oreohelix soccorensis soccorensis Pilsbry (fossil only) 



SYSTEMATIC DISCUSSION 

Pelecypoda 

Family Sphaeriidae 

Pisidium casertanum (Poli) 



General Distribution. — Worldwide; this is by far the most common Pisidium. 
Guadalupe Park Distribution. — Springs at base of Pipeline Canyon (Cheatum et al. 1972). 
Comments.— King (1948:145) reported a fossil Pisidium sp. from Bell Canyon, 1 mi. N of the 
Hegler Ranch. This, in addition to those in Cheatum et al. (1972), constitutes the only pelecypod 
reports for the Guadalupe Mountains. 

Gastropoda (freshwater) 
Family Hydrobiidae 



96 FULLINGTON 

Hydrobia sp. 
Comments. — A single fragmented shell was obtained from under a rock, in a pine stand, on the 
ridge (7400 ft) that forms the east side of Upper Dog Canyon. The location of this gill-breathing 
aquatic genus remains the greatest mystery of the Guadalupean snail fauna. 
Family Lymnaeidae 

Lymnaea humilis (Say) 

Fossoria obrussa (Say), in King (1948:145) as a fossil from Bell Canyon. 

Lymnaea (Galba) obrussa (Say), in Bequaert (1966) citing King (1948:145). 
General Distribution. — Canada, United States, and northern Mexico. 
Guadalupe Park Distribution. — Manzanita Spring (Cheatum et al. 1972) Choza Spring. 
Comments. — This is a relatively common lymnaeid in the Southwest and should be found in 
more aquatic localities in the Guadalupes. 
Family Physidae 

Physa virgata (Gould) 

Physa anatina Lea, in King (1948:145) Bell Canyon as a fossil. 

Physa forsheyi Lea, in Bequaert (1966), "living in pools and slowly running water of Bell 
Canyon, near the crossing of Hwy 180." 

General Distribution. — Common over most of south-central and southwestern North America. 
In California at least as far north as Sacramento and San Francisco Bay, northernmost Baja 
California, and eastward, southernmost Nevada, Utah, and Colorado; northward it is replaced 
by P. gyrina; southern Kansas (Taylor 1966:211). 

Guadalupe Park Distribution. — McKittrick Canyon, spring at base of Pipeline Canyon, drift in 
Delaware Creek. 

Comments. — This is probably the most common freshwater snail in the Southwest. 
Family Planorbidae 

Helisoma trivolvis thvolvis (Say) 

Helisoma trivolvis lentum (Say), in Cheatum et al. (1972) in Manzanita Spring. 
General Distribution. — "Atlantic coast and Mississippi River drainages, northward to Arctic 
British America and Alaska and southward to Tennessee and Missouri. The southern distribu- 
tion is not clear owing to the mixing with related species" (Baker 1928:332). 
Guadalupe Park Distribution. — Manzanita Spring, Choza Spring. 

Comments. — A conservative taxonomic approach is taken here in the relationship of H. t. 
trivolvis and H. t. lentum until the distribution and validity of lentum is established. In Texas, 
the shell characters are quite variable within populations and unreliable for taxonomic 
purposes. 

Gyraulus parvus (Say) 
General Distribution. — "Eastern North America, east of the Rocky Mountains from Florida to 
Alaska" (Baker 1928:377). 

Guadalupe Park Distribution. — Manzanita Spring (Cheatum et al. 1972). 
Comments. — The flat, open pond constructed at Manzanita Spring would appear to be a pre- 
ferred stopping place for migrating aquatic birds as indicated by the presence of H. t. trivolvis 
and G. parvus. Both are known to be easily transported in mud on the feet of such birds. 

Gastropoda (land) 

Family Helminthoglyptidae 

Humboldtiana ultima (Pilsbry) 
General Distribution.— Guadalupe Mountains: Pine Springs Canyon (Pilsbry 1939:408-409). 
Pilsbry and Ferriss during their trip in 1922 to the mountains on 2 November, went up the PX 
trail, apparently, to the area just north of Bush Mountain (Sta. 237). There they collected the 
first "Lysinoe." However, on 9 to 11 November, they collected living specimens "just above the 
box above the Gateway" (Sta. 240) (Devils' Hall). Apparently, Pilsbry described the holotype 
from this site. Also occurs in the Sierra Diablo Mountains (Cheatum and Fullington, unpubl.). 
Guadalupe Park Distribution.— General throughout the mountains above 6500 ft elevation. 
Usual habitat, forested ravines and rocky outcrops. 



MOLLUSCA 97 

Comments.— The shell in the northern portion of the mountains is larger (average height, 27.9 
mm), darker, and the bands are wider and much darker than the shell of the southern area 
(south of Lost Peak), where it becomes smaller (average height, 25.3 mm) and lighter. In the 
Bush Mountain area, the banding is highly variable. In the Aspen Grove area, most of the 
sampled population had only two bands rather than the usual three. 

King ( 1948: 145) reported H. ultima as a fossil from alluvial deposits on the north side of Pine 
Springs Canyon, 1 mi. W of Pine Springs. The thin shell does not lend itself well to fossilization 
although 1 do have fossil specimens from Sloth Cave. These specimens were sent to me by Lloyd 
Logan who is investigating the caves on the west side. 

The Guadalupe Mountains are the northernmost localities for this Mexican genus although 
other species occur on mountaintops in West Texas. It is also the largest snail indigenous to 
North America. 

Family Urocoptidae 

Metastoma {Holospira) roemeri roemeri (Pfeiffer) 
General Distribution. — Westward in Texas from the type locality at New Braunfels, Texas, to 
the Franklin Mountains at El Paso, Texas. It also occurs in the Sacramento and San Andreas 
mountains of southeastern New Mexico (Bequaert 1966). 

Guadalupe Park Distribution.— At elevations from 5000 to 6500 ft Pilsbry ( 1946: 1 1 5) reported 
it from the Guadalupe Mountains east of Orange, in Lincoln and Eddy counties, New Mexico, 
collected during the Pilsbry and Ferriss trip in 1922. It is almost certain that these localities are 
from the northwestern side of theGuadalupes, Culberson County, Texas. Pilsbry and Cheatum 
(1951:88) reported that a Mrs. Fischer collected it from Bone Canyon. The Dallas Museum of 
Natural History has collected it from the following localities: Bone Canyon, Pine Springs 
Canyon, Smith Spring Canyon, McKittrick Canyon, and Pipeline Canyon, 8100 ft. 
Comments. — M. r. roemeri apparently occurs only along the low-level, xeric rocky exposures 
that flank the mountains. It has not been collected within the mountains. Fossil specimens were 
taken from Sloth Cave. 

Holospira montivaga montivaga Pilsbry, 1946 
General Distribution. — Guadalupe Mountains: Type Locality, "New Mexico, Guadalupe 
Mountains east of Orange, the types from a terraced butte in a deep, dry canyon (our station 
240), about 2 miles south of the PX trail over the Mountains" (Pilsbry 1946: 124). Again, Pilsbry 
was in error thinking that he and Ferriss, during their 1922 trip, were in New Mexico. My esti- 
mate is that the site is near the north end of Blue Ridge as stated by Metcalf (1970:35). 
Guadalupe Park Distribution. — General over the mountains at elevations over 5000 ft. 
Comments. — As with Humboldtiana ultima, the shells of H. m. montivaga are larger (average 
height, 18.0 mm) in the northern section of the mountains and become smaller southward. The 
H. m. montivaga, H. m. form brevaria, H. pit vis, and H. oritis complex have become extremely 
problematic with respect to the data compiled during this survey. For the present time, all 
species are being retained, but evidence strongly leans toward a synonomization. The problem 
will be discussed later. 

Holospira montivaga form brevaria Pilsbry, 1946 
General Distribution. — Guadalupe Mountains: Type Locality, New Mexico, "eastern slope of 
Guadalupe Mountains near south end, above Walter Glover's ranch house in Pine Springs Can- 
yon, below the rock Gateway (Devil's Hall), (Pilsbry and Ferriss)" (Pilsbry 1946:125). 
Guadalupe Park Distribution. — General over the mountains at elevations over 6300 ft except in 
McKittrick Canyon. This distribution is based upon the strict definition of this form as having 
coarser ribbing than H. m. montivaga. Spire height differences (H. m. brevaria - 8-12 mm, H. 
m. montivaga - 15 mm; Pilsbry 1946) had no meaning when reasonably large populations were 
examined. 

Comments. — The only area where at least 85% of the population (46 specimens) sampled con- 
formed to this form was in the vicinity of the type locality. 

Holospira pityis Pilsbry and Cheatum, 1951 
General Distribution.— Guadalupe Mountains: Type Locality, "Guadalupe Range: Pine 
Springs Camp, Texas" (Pilsbry and Cheatum 1951). 



98 FULLINGTON 

Guadalupe Park Distribution. — Pine Springs Canyon, The Bowl near Pipeline Canyon, and 
the vicinity of Guadalupe Peak. 

Comments. -By description, H. pityis is distinguished by its very fine, close-set ribbing and 
height (8.5 mm) from H. m. brevaria. In The Bowl area near the eastern escarpment and Pipe- 
line Canyon, and in lower Pine Springs Canyon, the populations conform 90% to the descrip- 
tion. However, in the Guadalupe Peak area and upper (above 7000 ft) Pine Springs Canyon, the 
height and ribbing characters merge with the characters for H. m. form brevaria. Fossil 
specimens from Sloth Cave conform more closely to H. pityis, but the point could well be 
argued. 

Holospira oritis Pilsbry and Cheatum, 1951 
General Distribution. —Guadalupe Mountains: Type Locality, "this species was collected near 
Pratt's lodge in McKittrick Canyon in the Guadalupe Range approximately 7 miles from Pine 
Springs, Texas. They were on rocks of a juttingescarpmentof limestone near a stream" (Pilsbry 
and Cheatum 1951:90). 

Guadalupe Park Distribution. Apparently limited to McKittrick Canyon. 
Comments. — H. oritis differs (by type definition) from H. m. montivaga in being more robust, 
longer (height, 14 to 20 mm), and with one or more whorls being smooth. In one large popula- 
tion (in McKittrick Canyon proper) over 40% of the specimens conformed to H. m. montivaga. 
The oritis characters showed up in 10 % of a population of H. m. montivaga located in the 
canyon, with the spring forming the south end of Upper Dog Canyon. 

Holospira danielsi Pilsbry and Ferriss, 1915 
General Distribution. — Dragoon Mountains of Arizona. 
Guadalupe Park Distribution. — McKittrick Canyon. 

Comments. —One specimen was taken from drift at the mouth of McKittrick Canyon by 
Cheatum and Fullingtonin 1970 and was reported in Cheatum etal.( 1972). I have always been a 
little dubious of this specimen due to the distance from the type locality and because there is but 
one specimen. However, the species may actually occur in the Guadalupes and no one has dis- 
covered the population. Pilsbry (1946:137) cites the habitat as ". . . they live on the most 
exposed, hottest slopes, often in great profusion. . . ." He also states that they live under dead 
agaves, stones, and sotols. No one, as yet, has investigated this type of habitat in the Guadalupe 
Mountains. 

Due to the extent of this survey, which has been the most thorough of all molluscan investi- 
gations in the Guadalupe Mountains thus far, several statements may be made concerning the 
Holospira species complex or at least the problem itself may be better defined. 

Suspicion as to the validity of such species as H. montivaga form brevaria, H. pityis, and H. 
oritis have been voiced by several authors (Bequaert 1966; Cheatum and Fullington 1973:40, 
Bequaert and Miller 1973: 142). With this in mind, emphasis during this investigation was placed 
on large population analysis rather than upon simple species collecting by a few individuals. 
Unfortunately, time has not permitted necessary anatomical analysis. 

In all examined populations of Holospira m. montivaga, the range of shell characters 
(height, number, and coarseness of ribs) included that of H. m. form brevaria and vice versa. 
Several characters (some smooth whorls and height of at least 18 mm) for H. oritis were also 
present in a few H. m. montivaga populations outside McKittrick Canyon and vice versa. As 
stated earlier, H. m. montivaga decreases in height southward until, in the Pine Springs 
region, it becomes almost indistinguishable from brevaria except for the finer striatjons . 
Populations of //. pityis are much more restricted from The Bowl southward. McKittrick 
Canyon and Pine Springs Canyon appear to be "melting pots" for the characteristics which 
are used to distinguish species of Holospira. Thus, the problem becomes more question of 
which environmental pressures are operating on gene expression in the Guadalupe Holospira. 
The limited area of the Guadalupe Mountains should make them an excellent "laboratory" 
to further examine this problem. 

Excluding H. danielsi, all Guadalupean Holospira species probably should be synonymized 
under the earliest named H. m. montivaga. The separate species have been retained in this 
study for locality retention purposes in order to assist future investigations on this complex. 



MOLLUSCA 99 



Family Bulimulidae 

Rabdutus {Bulimulus) dealbaius dealbatus (Say) 
Bulimulus dealbatus pecosensis Pilsbry and Ferriss, in King (1948:145). 
Bulimulus dealbatus neomexicanus Pilsbry, in Bequaert (1966) and Cheatum et al. 
(1972:8). 

General Distribution. — Northern Mexico east of Sierra Madre Oriental; northward through 
Texas to eastern Oklahoma and Kansas; southwestern Missouri eastward to Alabama; dis- 
junct populations in the San Andreas, Sacramento, and Guadalupe mountains of New Mexico 
and Texas (Fullington and Pratt 1974:16). 

Guadalupe Park Distribution. — Common over the mountains from 5000 ft to over 8000 ft 
elevation; found in almost every type of habitat from dry, barren exposure to the high forested 
ravines. Pilsbry (1946:13) mentions that he and Ferriss collected it in 1922. King (1948:145) 
reported it being high up on the south wall of Pine Springs Canyon. Bequaert (1966) examined 
specimens collected by Monroe L. Walton in 1952 at the ruins of the old Butteriield Stage 
Coach Station, Pine Springs. Bequaert also found it living in South McKittrick Canyon. Field 
parties from the Dallas Museum of Natural History found it at almost every collecting site. 
Comments. — R. dealbatus neomexicanus is considered a large R. d. dealbatus (Fullington and 
Pratt 1974:16). 

Rabdotus {Bulimulus) dealbatus durangoanus (Martens, 1893) 
Bulimulus dealbatus pasonis Pilsbry, in Bequaert (1966). 
Rabdotus dealbatus durangoanus (Martens) in Fullington and Pratt (1974). 
General Distribution. — From the Sacramento Mountains, New Mexico, south through Trans- 
Pecos Texas into Mexico (Fullington and Pratt 1974). 

Guadalupe Park Distribution. — Pilsbry (1946:19), accompanied by C. Harvey in 1935, col- 
lected this species from the southern flank of Signal Peak (old name for Guadalupe Peak), south 
end of the Guadalupe Mountains, 6500 ft. The Dallas Museum of Natural History collected it in 
Bone Canyon, elevation 5300 ft. 

Comments. — This small form of R. dealbatus (as also suggested by Bequaert 1966) is probably 
more common around the flanks of the Guadalupes than records indicate. It generally lives 
under dead sotol and arborescent yuccas on arid, barren low slopes. Fullington and Pratt 
(1974:16) state that it occurs sympatrically with R. d. dealbatus in the Guadalupes. 

Family Thysanophoridae (Sagdidae) 

Thysanophora hornii (Gabb) 
General Distribution. — Northern Mexico, north to southeastern Arizona, southern New 
Mexico, and southwestern Texas (Fullington and Pratt 1974:27). 

Guadalupe Park Distribution. — King (1948:145) reported this species as a fossil from a clay 
deposit in Bell Canyon. Bequaert ( 1966) took it from drift in Bell Canyon and found it living in 
McKittrick Canyon. The Dallas Museum of Natural History collected it at Smith Spring, 
elevation 6000 ft. 

Comments. — This is not a common species and generally prefers dry, hot rocky areas although 
Fullington and Pratt (1974:27) state that in the mountains of Trans-Pecos Texas, it inhabits 
Pinyon-Oak-Juniper Woodland at higher elevations. A related western species, Microphysula 
ingersolli (Bland), should occur in the Guadalupes but has not been collected thus far. 

Family Endodontidae 

Helicodiseus nummus (Vanatta) 
General Distribution. — Texas, from the Balcones escarpment southward and westward to the 
Guadalupe Mountains; one locality in Indiana (Pilsbry 1946:639). 

Guadalupe Park Distribution. — Sparsely distributed from canyons at south end of Upper Dog 
Canyon (6500 ft) to the Aspen Grove (7750 ft). It has not been previously reported from the 
Guadalupe Mountains. 

Comments. — The Helicodiseus complex has become almost bewildering with the recent addi- 
tion of several new species and the elevation of several subspecies to specific rank by various 
authors. This tiny species (diameter, 1.3 to 1.5 mm) is difficult to separate from other immature 



100 FULLINGTON 

smooth-shelled Helicodiscus . It is interesting that H. nwnmus occurs only at higher elevations 
in the Guadalupes, whereas elsewhere in Texas, it appears to be a lower, moist, woodland form. 

Helicodiscus singleyanus singleyanus (Pilsbry) 
General Distribution. — New Jersey to South Dakota, Colorado, Arizona, New Mexico, and 
Texas; also Florida to Louisiana (Pilsbry 1948:636). 

Guadalupe Park Distribution. — Bequaert (1966) collected the first reported specimen in South 
McKittrick Canyon (5300 ft). The Dallas Museum of Natural History collected it from the can- 
yons at the south end of U pper Dog Canyon (6300 to 6500 ft). I also have identified several fossil 
specimens from Sloth Cave. 
Comments. — This is a common snail in Texas. 

Helicodiscus eigenmanni Pilsbry, 1890 
Helicodiscus arizonensis (H. eigenmanni arizonensis) Pilsbry and Ferriss, in Cheatum et 
al. (1972). 

General Distribution (includes the distribution of H. e. arizonensis). — South Dakota to 
Arizona, New Mexico, Texas, and Mexico (Pilsbry 1948:630). 

Guadalupe Park Distribution. — Drift in McKittrick Canyon, otherwise sparsely distributed 
over the mountains in the wooded ravines only at elevations of 6500 to over 8000 ft. Dallas 
Museum of Natural History localities include Pipeline Canyon, Smith Spring Canyon, near top 
of El Capitan, Aspen Grove, Upper Dog Canyon area, Blue Ridge area. 
Comments. — Bequaert and Miller (1973:85-87) also have relegated H. e. arizonensis to 
synonomy. They added that H. eigenmanni might be only a southwestern subspecies of H. 
parallelus. 

Discus cronkhitei (Newcomb 1865) 
General Distribution.— Canada and most of the United States (Pilsbry 1948:602-603). 
Guadalupe Park Distribution. — As a fossil, north side of Pine Springs Canyon, 1 mi. W Pine 
Springs (King 1948:145), McKittrick Canyon drift (Cheatum etal. 1972), and Sloth Cave on the 
west side. Field parties from the Dallas Museum of Natural History found it living from Upper 
Dog Canyon to The Bowl and Blue Ridge at elevations over 6500 ft. 

Comments. — Although this species is a relatively common fossil across Texas, this is the first 
report of this species as Recent for the state. At least six sampled populations contained speci- 
mens with smooth bases and slightly more elevated spires conforming to Discus shmekii 
(Pilsbry). This situation apparently is common as other malacologists have observed it as well. 

Punctum minutissimum (Lea) 
General Distribution.— Maine to Florida, west to Oregon and New Mexico (Burch 1962:80). 
Guadalupe Park Distribution. — The Dallas Museum of Natural History collected this tiny 
species in densely wooded ravines from Upper Dog Canyon southward to the Aspen Grove but 
only at elevations over 6600 ft. 

Comments. — This is the first report of this species for West Texas either as Recent or fossil. Mr. 
Lloyd Pratt (pers. comm.) has collected it alive in Tarrant County, Texas. 

Punctum vitreum H. B. Baker 1930 
General Distribution. — From New Jersey through most of the southeastern states to north- 
eastern Mexico (Baquaert 1966). 

Guadalupe Park Distribution. — Bequaert (1966) reported this species living in South 
McKittrick Canyon. 

Comments.— The museum has not collected P. vitreum in the park. Because most of our work 
has been concentrated at higher elevations where P. minutissimum occurs, it may be that P. 
vitreum inhabits the lower levels. 
Family Zonitidae 

Zonitoides arboreus (Say, 1816) 
General Distribution. — North American continent to Costa Rica. 

Guadalupe Park Distribution.— As a fossil in Pine Springs Canyon, and at the east side of Bell 
Canyon (King 1948: 145). The Dallas Museum of Natural History found living specimens gen- 
erally distributed at elevations above 6300 ft. Cheatum et al. (1972) reported it as Recent for 



MOLLUSCA 101 

Culberson County. The localities were McKittrick Canyon, Pine Springs Canyon, Smith 
Spring Canyon, and The Bowl. 

Comments. — The probable reason that Z. arboreus was not reported as living in the Texas 
Guadalupes until 1972 is that until the museum began investigating the mountains in 1969, no 
one, except Pilsbry and Ferriss in 1922, collected higher than the base of the mountains. 
Glyphyalinia indentata paucilirata (Morelet, 1851) 
Glyphyalinia indentata (Say), in Cheatum et al. (1972). 
General Distribution.^Central and southern United States southward to Guatemala (Pilsbry 
1946:291). 

Guadalupe Park Distribution. — Very general over the mountains at all elevations. King 
( 1948: 145) reported it as a fossil in Pine Springs Canyon. Bequaert ( 1966) collected washed up 
specimens in Bell Canyon in 1958 and live in McKittrick Canyon. The Dallas Museum of 
Natural History found it at most collecting sites. 

Comments. — Cheatum et al. ( 1972) placed G. i. paucilirata as a synonym under G. i. indentata. 
G. i. paucilirata is characterized externally by a much wider umbilicus. The Guadalupe shells 
conform quite well to this character and, thus, the subspecies is retained here. However, as with 
several other gastropod species, G. i. paucilirata may prove to be a western form of G. i. 
indentata. 

Nesovitrea sp. 
Comments.— A small zonitid was collected at several localities. It compares in external form 
with Nesovitrea binneyana occidentalis(H. B. Baker), and is the same type of shells identified by 
Cheatum et al. (1972) as Glyphyalinia roemeri (Pilsbry and Ferriss). Until this series has been 
more carefully examined, it is best to leave it unnamed for the present. 

Euconulus fulvus (Muller, 1974) 
General Distribution. — General throughout the cold temperate parts of Europe, Asia, and 
North America; in the southwestern United States, restricted to mountains above 5000 ft 
(Bequaert 1966). 

Guadalupe Park Distribution. — Bequaert (1966) first recorded it as living in South McKittrick 
Canyon. The Dallas Museum of Natural History has found it to be distributed generally over 
the mountains at elevations above 6300 ft. 

Comments.— E. fulvus is a fairly common fossil in Texas and as far as is presently known, it is 
extant within the state only in the Guadalupe and Chisos mountains (Cheatum et al. 1972). 

Striatura meridionalis (Pilsbry and Ferriss, 1906) 
General Distribution. — Widespread in the southeastern United States, northward to New 
Jersey, westward to Arizona, and southward into northwestern Mexico (Bequaert 1966). 
Guadalupe Park Distribution. — Bequaert (1966) was the first to report this species in the 
Guadalupe Mountains. His specimens came from South McKittrick Canyon. The Dallas 
Museum of Natural History collected it over most of the mountains at elevations above 6500 ft. 
Comments.— Cheatum et al. (1972) reported it as a fossil from the Guadalupes. It is always 
found in the high, densely wooded canyons under humus. 

Vitrina pellucida alaskana Dall, 1905 
General Distribution. — Alaska to California, Arizona and New Mexico, eastward to South 
Dakota (Bequaert and Miller 1973:73-74). 

Guadalupe Park Distribution. — Very sparsely distributed over the mountains at elevations 
over 7000 ft. 

Comments.— This is the first report of V. p. alaskana for Texas and the Guadalupe Mountains. 
It was collected at most of our sites but only in low numbers. Bequaert and Miller (1973:73-74) 
placed V. alaskana Dall (Pilsbry 1946) as a subspecies of V. pellucida (O. F. Muller 1774) of the 
Old World. 

Hawaiia minuscula minuscula (Binney) 
General Distribution.— North American Continent. 

Guadalupe Park Distribution.— General over the mountains at elevations above 6200 ft. 
Comments.— Cheatum et al. (1972), without realizing, gave the first report of this very com- 



102 FULLINGTON 

mon species for the Guadalupes. Since that report, we have found it to be common at higher 
elevations and as a fossil from Sloth Cave. 

Hawaiia minuscula neomexicana (Cockerell and Pilsbry, 1900) 
General Distribution. — Scattered localities in Texas, New Mexico, and Mexico. 
Guadalupe Park Distribution. — King ( 1948: 145) reported this species as a fossil from the north 
side of Pine Springs Canyon. Bequaert( 1966) found it in drift from Bell Canyon at Highway 180 
and living in South McKittrick. I found it living at scattered localities across the mountains at 
elevation above 6200 ft and as a common fossil from Sloth Cave. 

Comments. — Bequaert (1966) feels that this subspecies may be just a slight variance of H. m. 
minuscula. However, in a mixed lot, they are easily separable. 
Nesovitrea electrina (Gould, 1841) 
General Distribution. — Alaska, through Canada to Virginia, south to New Mexico and north- 
eastern Arizona (Bequaert and Miller 1973:67-68). 

Guadalupe Park Distribution. — Spottily distributed over the mountains at elevations over 6300 
ft. This is the first report of this species as living in Texas, although it is a common fossil. 
Comments. — I am not following Bequaert and Miller (1973:67-68) in placing electrina 
(elevated by Pilsbry 1946:256 to specific rank) as a subspecies of the Holarctic N. hammonis at 
this time. 

Family Limacidae 

Deroceras laeve (Miiller, 1774) 
General Distribution. — Arctic Region to Central America (Bequaert and Miller 1973:68-69). 
Guadalupe Park Distribution. — Upper Pine Springs Canyon and Choza Spring. 
Comments. — This is the first report of a slug in the Guadalupe Mountains. 

Family Polygyridae 

Ashmunella rhyssa rhyssa (Dall) 
General Distribution. — Living specimens known from the Sacramento Mountains, New 
Mexico (Metcalf and Fullington, in litt.). 

Guadalupe Park Distribution. — As a fossil from Pine Springs Canyon. 
Comments. — Dr. A. L. Metcalf informed me that he recently collected this species in an alluvial 
deposit where he also collected a new, very small Ashmunella that we are currently describing. 

Ashmunella kochi amblya Pilsbry, 1940 
General Distribution. — Guadalupe Mountains: Type Locality, "Guadalupe Mountains, east 
and southeast of Orange, New Mexico, from Pine Springs Canyon, above Walter Glover's 
House (Pilsbry and Ferriss, November 2-11, 1922)" (Pilsbry 1940:977). Bequaert (1966) stated 
that in the Harvard Museum of Comparative Zoology there are paratypes (of the original lot, 
received from Pilsbry) labeled more precisely, "Canyon south of the summit of PX trail, east of 
Orange." However, no altitude was given. Also found in western foothills of San Andreas 
Mountains and Organ Mountains, New Mexico (Bequaert 1966). 

Guadalupe Park Distribution. — General over the mountains at elevations over 7000 ft. Mainly 
in the wooded canyons but almost as readily found in the craggy limestone outcrops nearly 
devoid of vegetation. King (1948:145) reported it as living on the upper south wall of Pine 
Springs Canyon and as a fossil from the north side of Pine Springs Canyon. M. L. Walton 
(1963:126) also recorded it as living Wi mi. up Pine Springs Canyon. 

Comments.—/!, kochi amblya is distributed in the Texas Guadalupes from Guadalupe Peak 
north to the ridge that transversely bisects the mountains which includes Lost Peak and the 
north side of McKittrick Ridge. Spire height, shell diameter, and teeth structure are highly 
variable. Shells from the Blue Ridge area populations conform perfectly to A. kochi kochi 
Clapp which is known from only one locality — Black Mountain, at the south end of San 
Andreas Mountains at 6800 ft, Dona Ana County, New Mexico (Pilsbry 1940:979). A. kochi 
amblya should probably be synonomized under A. kochi kochi. 

Ashmunella edithae Pilsbry and Cheatum, 1951 
General Distribution. -Texas Guadalupe Mountains: Type Locality "Guadalupe Range, near 
the top of the mountain up Pipeline Canyon, Pine Springs, Texas" (Pilsbry and Cheatum 
1951:88). 



MOLLUSCA 103 

Guadalupe Park Distribution. — The distribution of A. edithae in the mountains is unclear at 
present. 

Comments. -Ashmunella edithae should probably be synonomized with A. kochi amblya for 
several reasons. The original description by Pilsbry and Cheatum (1951:88) matches perfectly 
the description of A. carlsbadensis, as does their illustration. I have collected the entire length of 
Pipeline Canyon twice and have collected nothing but A. kochi amblya. This fact leads to the 
main problem — the type locality. Pilsbry and Cheatum cite the locality as "Pipeline Canyon." I 
have shells listed as paratypes from Cheatum's personal collection with the locality cited as 
"Guadalupe Mountains on trail to Guadalupe Springs at base of El Capitan Mountain" (E. P. 
Cheatum, June 1950). Cheatum and his wife Edith (who collected the shells) both have told me 
that she went up the trail to Guadalupe Peak and not Pipeline Canyon. The three shells listed as 
paratypes are perfect A. kochi amblya, which makes one suspicious of mixed labels. The illus- 
trations of A. kochi amblya and A. edithae in Cheatum and Fullington ( 1 97 1 ) are also remark- 
ably similar. 1 believe that Edith collected A. kochi amblya and that Dr. Pilsbry simply had a 
lapse concerning A. carlsbadensis. Besides, no one at that time had collected A. carlsbadensis 
previously in Texas. Dr. Cheatum indicated that he collected in McKittrick Canyon. It may be 
that he picked up carlsbadensis there and in the process of shipping the material to Dr. Pilsbry, 
mixed these shells and Edith's material from wherever she collected them. 

Ashmunella carlsbadensis Pilsbry, 1932 
General Distribution. — Guadalupe Mountains of New Mexico and Texas: Type Locality, "A 
cave in Dark Canyon, southwest of Carlsbad, from the surface to a depth of two feet (E. B. 
Howard)" Pilsbry (1940:978). 

Guadalupe Park Distribution. — This species has a strange distribution in the Texas Guada- 
lupes. The Dallas Museum of Natural History collected it from the heavily wooded canyons of 
Upper Dog Canyon (7400 ft), and along the eastern flanks from McKittrick Canyon to Bone 
Canyon (5400 ft) on the west, except in the Pine Springs Canyon area. While A. kochi amblya 
occupies the higlf central areas, A. carlsbadensis encircles the mountains, inhabiting the lower 
level, rocky outcroppings. It appears to be tolerant of xeric conditions, whereas A. kochi 
amblya is not. Altitude is not a prohibitive factor as we found it in the limestone outcroppings 
along the summit of the eastern escarpment (8200 ft). It was found sympatrically with A. kochi 
amblya at only one locality — near the junction of Pipeline Canyon and The Bowl where the 
forest borders the edge of the escarpment. 

Comments. — The relationship of A. kochi amblya and A. carlsbadensis is interesting in that the 
two are sympatric at only one place (thus far investigated) in the mountains. The problem needs 
further investigating before a definitive statement may be made. In 1972, the late Dr. Cheatum 
and I collected a series of small Ashmunella along the east-facing Cliffs of the Sierra Diablos 
which lie just south of the Guadalupes that, cursorily, appear to be small carlsbadensis, but 
some evidence indicates that they may be a new species. 

In view of the fossil Ashmunella collected by Dr. A. L. Metcalf, and the unknown species 
from Sloth Cave, the whole Guadalupe Ashmunella complex appears to be highly dynamic. 

Ashmunella spp. 
Comments. — Two small species of fossil Ashmunella have been collected recently in the Guada- 
lupes — one by Dr. A. L. Metcalf from Pine Springs Canyon which we are in the process of 
describing and the second was taken from Sloth Cave but cannot be described until whole speci- 
mens are exhumed. 

Family Succineidae 

Succinea luteola Gould, 1848 
General Distribution. — Louisiana, Texas, New Mexico, Arizona, and Mexico (Pilsbry 
1948:829). 

Guadalupe Park Distribution. — King (1948:145) reported this species as a fossil from Pine 
Springs Canyon and from Bell Canyon. The Dallas Museum of Natural History collected 
throughout the mountains dead shells that conform to S. luteola, but no live specimens were 
taken. 



104 FULLINGTON 

Comments. — Due to the current state of confusion over succineid taxonomy, a conservative 
approach to this group has been taken in this report. 

Succinea sp. 
Comments. — An unusual series of fossil succineids were taken from Sloth Cave. These require 
further study and will be described at a later date. 

Family Cionellidae 

Cionella lubricella (Porro) 
Cionella lubrica (Muller 1774), in Bequaert (1966). 
General Distribution.— A worldwide genus with all North American forms referred to as 
Cionella lubrica (Pilsbry 1948:1047). Current consensus is that at least five species occur in 
North America (including Mexico) (F. W. Grimm, 1974 in litt., and Leslie Hubricht 1974 pers. 
comm.), and that the range of C. lubricella covers the area of this report. 
Guadalupe Park Distribution. — General over the mountains at all elevations where enough 
vegetation is present to maintain a layer of leaf litter. 

Comments. — Bequaert (1966) first reported this species from the Guadalupe Mountains as a 
possible fossil from drift in Bell Canyon. Cheatum et al. (1972) reported it living in the Guada- 
lupes. The Dallas Museum of Natural History found it to be one of the most common species in 
the park. Bequaert and Miller (1973:72) reported it living in McKittrick Canyon (as Cochlicopa 
lubrica). 

Family Pupillidae 

Gastrocopta armifera (Say, 1821) 
General Distribution. — Canada, eastern United States to Colorado, New Mexico, and Texas 
(Cheatum and Fullington 1973:10). 

Guadalupe Park Distribution. — Bequaert (1966) first collected this species from drift in Bell 
Canyon at Highway 180 as a fossil; Cheatum et al. (1972) reported it living in McKittrick Can- 
yon. Metcalf (1975, pers. comm.) found it as a fossil in Pine Springs Canyon. 
Comments. — Field parties of the Dallas Museum of Natural History did not find G. armifera 
alive during the July investigation, thus its distribution in the park remains unknown. Evidence 
indicates that this species is losing ground in the West. Bequaert and Miller (1973: 172) indicate 
that it no longer lives in Arizona, but that there are numerous fossil localities. This situation also 
holds true for New Mexico and Texas. 

Gastrocopta procera procer a (Gould, 1840) 
General Distribution. — Eastern United States, west to the Dakotas, south to the Rio Grande, 
with only spotty records in New Mexico and Arizona. 

Guadalupe Park Distribution. — One colony was found by the museum (July 1974) in one of the 
head canyons of Upper Dog Canyon, elevation 6200 ft. 

Comments. — This is the first report of G. p. procera in the park. In the western United States, it 
is not considered a native snail by Bequaert and Miller (1973:90-91). 

Gastrocopta pentodon (Say, 1821) 
General Distribution. — Northeastern Canada, the eastern half of the United States (as far west 
as Colorado), and northeastern Mexico (Bequaert 1966). 

Guadalupe Park Distribution.— Bequaert (1966) reported the first specimens from McKittrick 
Canyon. Cheatum et al. (1972) also reported it from McKittrick Canyon. During July 1974, the 
Dallas Museum of Natural History found it spottily across the mountains at elevations 6200 to 
7500 ft. 

Comments.— Bequaert and Miller (1973:88-90) placed G. tappaniana (Adams) under the 
synonomy of G. pentodon. 

Gastrocopta pilsbryana (Sterki, 1890) 
General Distribution. — In the mountains of New Mexico, Arizona, Mexico, and one locality in 
southwestern Utah (Bequaert and Miller 1973:158-159), Franklin Mountains (Metcalf and 
Johnson 1971:98), and Guadalupe Mountains, Texas (Cheatum et al. 1972), as fossils. 
Guadalupe Park Distribution.— Bequaert (1966) presented the first report for the park and for 
Texas wher he found it living in South McKittrick Canyon. Cheatum et al. (1972) reported it as 



MOLLUSCA 105 

a fossil from the Guadalupes. The Dallas Museum of Natural History found it to be common 
across the park at elevations over 6500 ft. 

Comments. — At present, the Guadalupe Mountains are the easternmost limit for this western 
mountain-dwelling species. 

Gastrocopta ashmuni (Sterki, 1898) 
General Distribution. — Arizona, New Mexico, northern Mexico, and Trans- Pecos Texas. 
Guadalupe Park Distribution.— Mead in 1969 {in Bequaert and Miller 1973) reported it from 
South McKittrick Canyon as did Cheatum et al. ( 1972). The Dallas Museum of Natural History 
also found it only in McKittrick Canyon. Fossil specimens were taken from Sloth Cave. 
Comments.— Generally, this species lives in protected areas of low, rocky hills, usually at eleva- 
tions between 3000 and 6000 ft. This type of habitat has not been fully investigated in the Guada- 
lupes. When the area has been more fully covered, G. ashmuni will probably turn up in more 
places than just McKittrick Canyon. 

Gastrocopta contracta (Say, 1822) 
General Distribution. — Southeastern Canada, widespread over most of the eastern United 
States, west to the Dakotas, south to northern Mexico; in the Southwest, Trans-Pecos Texas 
and eastern New Mexico. 

Guadalupe Park Distribution. — Bequaert (1966) first reported it from flood debris in Bell 
Canyon in 1958. Later, he found it living in South McKittrick Canyon. The Dallas Museum of 
Natural History found it distributed across the mountains at elevations above 6300 ft. 
Comments. — Bequaert and Miller (1973:91) consider G. contracta to be a post-Pleistocene 
invader of the Southwest. 

Gastrocopta pellucida (Pfeiffer, 1841) 
Gastrocopta pellucida hordeacella (Pilsbry 1890), in Bequaert (1966). 
General Distribution.— A common snail in the southern United States from New Jersey to 
southern California; southward to Central America (Bequaert 1966). 
Guadalupe Park Distribution. — In 1958, Bequaert collected the first Guadalupe specimens 
from drift in Bell Canyon. Later, he took live specimens from South McKittrick Canyon, eleva- 
tion 5300 ft. We found only two colonies, one at The Bowl (8200 ft) and the other at Smith 
Spring (6000 ft). 

Comments.— Bequaert and Miller (1973:79-81) synonomized G. pellucida hordeacella and G. 
parvidens under G. pellucida. 

Pupilla muscorum muscorum (Linne, 1758) 
General Distribution.— Widespread throughout temperate Europe, Asia, and North America; 
in North America as far south as the mountains of New Mexico (as Recent) (Bequaert 1966). 
Often found as a fossil in Texas. 

Guadalupe Park Distribution.— King (1948:145) has reported the only instance of this species 
in the park. It was found in a fossilferous layer on the north side of Pine Springs Canyon, 1 mi. 
W of Pine Springs, Texas. 

Comments. — P. m. muscorum is considered a mountain-dwelling snail in the west and probably 
will be found alive in the park. It and P. blandi apparently were eliminated from the Great 
Plains just before or after the Bradyon interval (Leonard and Frye 1962:27). 

Pupilla blandii (Morse, 1865) 
General Distribution.— Throughout the Rocky Mountains, from Alberta to Arizona and New 
Mexico (at high altitudes) according to Bequaert (1966). 

Guadalupe Park Distribution.— Bequaert (1966) first reported this species in the park from 
drift in Bell Canyon, which he considered fossil. Cheatum and Fullington (1973:23) collected 
shells in McKittrick Canyon drift with the dried animal still within the shell. The museum did 
not collect this species at the higher elevations. 
Comments. — See comments under Pupilla sonorana. 

Pupilla sonorana (Sterki, 1899) 
General Distribution. — New Mexico and west Texas. 
Guadalupe Park Distribution.— In 1969, Mead (in Bequaert and Miller 1973) first reported this 



106 FULLINGTON 

species from South McKittrick Canyon. The Dallas Museum of Natural History found it to be 
common over the entire park at elevations above 6300 ft. 

Comments. — Pupilla sonorana and P. blandii are probably the same species; the only dif- 
ference is in size and the compression behind the lip of sonorana. In most populations sampled, 
there were specimens that conformed to both species, except for size. As most fell into the height 
range (2.5 to 3.25 mm) of P. sonorana, for the present I am classifying them as P. sonorana. 
However, I also firmly feel (as do Bequaert and Miller 1973) that P. blandii and. P. sonorana are 
conspecific. 

Pupoides albilabris (Adams, 1841) 
General Distribution.— Southern Ontario; the United States westward to Colorado, Utah, and 
Arizona; northern Mexico, Cuba, Haiti, Puerto Rico, and Bermuda (Pilsbry 1948:921). 
Guadalupe Park Distribution. — In drift at Delaware Creek at State Highway 652 (Cheatum et 
al. 1972). 

Comments. — It is amazing that this ubiquitous gastropod has not been collected in the park 
more often than just the one record indicates. 

Pupoides hordaceus (Gabb, 1866) 
General Distribution. — Colorado, New Mexico, Arizona, Wyoming, and Utah (Bequaert and 
Miller 1973:58-59). It occurs only as a fossil in Texas and Kansas. 

Guadalupe Park Distribution. — Presumably as a fossil from drift at Delaware Creek at State 
Highway 652 crossing (Cheatum et al. 1972). 

Comments. — This species is yet to be collected as Recent in the Guadalupe Mountains. It pre- 
fers arid plateaus and xeric foothills and does not occur in the higher elevations according to 
Pilsbry (1948:924). This type of habitat in the park has not been fully explored as yet. 

Vertigo milium (Gould, 1840) 
General Distribution. — Recent from southern Ontario, south to Florida, west to Minnesota, 
South Dakota, northern Colorado, east Kansas to east-central Texas; only as a fossil westward 
through Arizona (with a few exceptions), no reports from New Mexico (Bequaert and Miller 
1973:95-95). 

Guadalupe Park Distribution. — Drift specimens from McKittrick Canyon (Cheatum et al. 
1972). 

Comments. — The specimens collected in McKittrick Canyon appear much too fresh to be 
classified as fossil, though they might well be. V. milium is another eastern Nearctic species 
that appears to be losing ground in the West. Areas such as the Guadalupe Mountains offer the 
last suitable habitats for this tiny snail in the West. 

Vertigo gouldii arizonensis Pilsbry and Vanatta, 1900 
General Distribution. — Arizona, New Mexico, and northern Mexico (Bequaert and Miller 
1973:186-187). 

Guadalupe Park Distribution.— Aspen Grove area to the northernmost edge of Blue Ridge at 
elevations over 7500 ft. 
Comments.— This is the first report of this subspecies for Texas and for the Guadalupes. 

Vertigo ovata (Say, 1 822) 
General Distribution. — Throughout most of temperate North America, becoming sporadic in 
the Rocky Mountain states (Bequaert and Miller 1973:92-94). 

Guadalupe Park Distribution.— Bequaert (1966) first reported V. ovata as a fossil from drift in 
Bell Canyon. Cheatum et al. (1972) reported it as Recent from drift in Delaware Creek at State 
Highway 652. I have reexamined the shells and they do appear very fresh, but they could very 
well be sub-Recent. 

Comments.— With further investigation, living material will most likely be found in the park. 
Family Valloniidae 

Vallonia perspectiva Sterki, 1893 
General Distribution.— Common throughout most of the southern United States; north to New 
Jersey and westward to North Dakota and Arizona (Bequaert and Miller 1973:96-98). 
Guadalupe ?ark Distribution.— Bequaert (1966) first collected this species as a fossil in South 



MOLLUSCA 107 



McKittrick Canyon at 5300 ft. Cheatum et al. (1972) reported it as Recent from drift in 
McKittrick Canyon. Bequaert and Miller (1973:26) stated that Miller in 1969 also found it living 
in South McKittrick Canyon. During July 1974, the Dallas Museum of Natural History found 
it commonly distributed throughout the park at elevations over 6300 ft in the forested ravines. 
Comments. — In Texas, V. perspectiva is known as living only in Terrell County, Chisos Moun- 
tains in Brewster County, and Franklin Mountains in El Paso County (Fullington and Pratt 
1974:29). 

Vallonia cyclophorella Sterki, 1893 
General Distribution.— A western montane snail from Washington to Arizona and New 
Mexico, a few relictal colonies in sheltered canyons along the eastern escarpment of the High 
Plains (Fullington and Pratt 1974:28). 

Guadalupe Park Distribution.— King (1948:145) reported it as a fossil from Pine Springs 
Canyon. 

Comments. — V. cyclophorella is another western species that apparently spread eastward 
during the Pleistocene and now is losing ground as the areas east of the Rockies become more 
arid. 

Vallonia gracilicosta Reinhardt, 1883 
General Distribution. — Rocky Mountain Region to New Mexico; only as a fossil in Arizona 
and previously in Texas. 

Guadalupe Park Distribution.— Cheatum et al. (1972) found fossil specimens in drift in Cherry 
Canyon at State Highway 180. Metcalf (pers. comm.) found it as a fossil in Pine Springs 
Canyon. The museum collected specimens from Upper Pine Springs Canyon (6600 ft) and along 
the Guadalupe Peak trail (7500 ft). 

Comments. — This is the first report of V. gracilicosta as Recent for both Texas and the park. 
The specimens from Upper Pine Springs Canyon could be questionable as being Recent or 
fossil, but the Guadalupe Peak shells are bright and fresh. 

Vallonia parvula Sterki, 1893 
General Distribution.— Southern Ontario to South Dakota, south to Iowa, Oklahoma, and the 
Texas Panhandle (Fullington and Pratt 1974:29). 

Guadalupe Park Distribution. — Fossil shells from drift in Delaware Creek at State Highway 
652 (Cheatum et al. 1972). 

Comments.— This species is listed with reservation. The shells from the only locality in the park 
have disappeared and cannot be reverified. They may have been young V. gracilicosta. 

Family Strobilopsidae 

Strobilops labyrinthica (Say, 1817) 
General Distribution. — Southern Maine and Quebec; south through Minnesota, Nebraska, 
and Kansas; east of the Great Plains to Oklahoma, Arkansas, Georgia, and Alabama (Fulling- 
ton and Pratt 1974:25). 

Guadalupe Park Distribution. — Living specimens were collected in McKittrick Canyon by 
Cheatum et al. (1972). 

Comments. — Only one other locality of living 5. labyrinthica is known in Texas (Harris 
County). The species is known from several Pleistocene sites in Texas. Thus it is another 
northern species that is becoming extinct in the Southwest. 

Family Oreohelicidae 

Oreohelix soccorensis soccorensis Pilsbry, 1905 
Oreohelix yavapai compactula Cockerell, in King (1948:145), from the north side of Pine 
Springs Canyon, 1 mi. W Pine Springs (fossil). 

General Distribution. — Known as Recent only in the Gallinas Mountains, New Mexico; as a 
fossil in the San Andreas and Sacramento mountains in New Mexico, and the Franklin, Hueco, 
and Guadalupe mountains in Texas (Metcalf, per. comm.). 

Guadalupe Park Distribution. — Metcalf (pers. comm.) has collected fossil shells in Pine 
Springs Canyon, in addition to King's (1948) report. This species was abundant in Sloth, Upper 
Sloth, and Dust caves. 



108 FULLINGTON 

Comments. I have examined one Oreo helix soccoren.u.s shell from the Guadalupe Peak area 
brought to me hy a group of young people without more information than the ahove. The shell 
looks very fresh. Although the outer surface shows evidence of weathering, the interior is com- 
pletely devoid of dirt and has only lungul myeclia in it. However, temerity and lack of better 
collecting site data force me to include it as a fossil. 

DISCUSSION 

Sixty-one species of land and freshwater mollusks are presently known 
from the Guadalupe Mountains National Park, Texas; 54 are Recent and 7 
are fossil forms. Doubtless, many more species will be discovered with fur- 
ther investigation. However, enough species are known now to enable us to 
understand something of the mountains' molluscan fauna. 

Apparently, the mountains have stood for a long time period as an 
"island" for periodic molluscan movements in relation to climatic changes. 
With the current southwestern drying trend, the mountains offer a last 
refuge for many of the more temperate species. It may be that even these 
"refuges" are losing species. The fossil record of the Guadalupes is still too 
poorly known for definite statements on this problem. 

Seventeen species in the Guadalupes currently have a more northerly or 
wide distribution (Pisidium casertanum, Lymnaea humilis, Helisoma 
trivolvis trivolvis, Gyraulus parvus, Helicodiscus singleyanus singleyanus, 
Discus cronkhitei, Zonitoides arboreus, Euconulus fulvus, Hawaiia mi- 
nuscula minuscula, Nesovitrea electrina, Deroceras laeve, Gastrocopta con- 
tractu, Pupilla muscorum muscorum, Pupoides albilabris, Vertigo milium, 
and Strohilops labyrinthica). Based on Southwestern fossil evidence, most 
of these species are rapidly losing ground in the area or are already known 
only as fossils (N. electrina, /). cronkhitei, S. labyrinthica, and E. fulvus). 

Twelve species inhabit only the Southwest (Physa virgata, Metastoma 
roemeri roemeri, Rabdotus dealbatus dealbatus, R. dealbatus durango- 
anuSt Thysanophora homti, Helicodiscus nummus, Glyphyalinia indentata 
pauciliratd, Hawaiia minuscula ncomexicana, Succinea luteola, and 
Clone I la lubricella). It is interesting to note that several southwestern 
genera {Pofygyra, Helicina, and Euglandina) apparently have never reach- 
ed the Guadalupes, while some almost strictly Mexican genera such as 
Htttnboldtiona and Thysanophora have made the transition. 

Species from the Rocky Mountain molluscan fauna, as defined by 
Bequaerl and Miller (1973:12), invaded the Guadalupes during cooler 
Pleistocene periods. Most oi these species (Holospira danielsi, Helicodiscus 
eigenmanni, (iastrocopta pilsbrvana, Gastrocopta ashmuni, Pupilla blandi, 
Pupilla SOnorana, Pupoides hordeaceus. Vallonia cyclophorella, Vallonia 
gracilicosta. \ itnna pellucida alaskana, and Oreohelix soccorensis soc- 
corensis) are found only at the higher elevations in the Guadalupes. An un- 
answered question is why several of these forms are now apparently extinct 
in the mountains ( Oreohelix and Pupoides hordeaceus) and why other forms 
such as Sottorella and Radiodiscus never became established. It should also 



MOLLUSCA 109 

be stated that forms such as Sonorella and Humboldtiana do not fossilize 
due to their thin shells and the elevations which they inhabit. 

The Guadalupes harbor only six species that inhabit the southeasterly to 
south-central United States (Striatura meridionalis, Gastrocopta armifera, 
Gastrocopta procera procera, Gastrocopta pentodon, Gastrocopta 
pellucida, and Vallonia perspectiva). The mountains are the westernmost 
limits for several of these species. 

Nine species are endemic, or nearly so, to the Guadalupe Mountains as 
follows: Humboldtiana ultima, Ashmunella carlsbadensis, Ashmunella 
kochi amblya, Ashmunella edithae, Holospira montivaga montivaga, H. 
montivaga form brevaria, H. pityis, H. oritis, and Ashmunella sp. nov. It 
should be understood, however, that, with further investigation, this number 
of species probably will be reduced through synonymy. 

All of the aquatic species represented are adventive and widespread in 
North America. Thus far, very few fossil aquatic species are recorded. 
Apparently, the stream systems of the Guadalupes have always had rela- 
tively sharp inclines with rapid run-off, and the Guadalupes never estab- 
lished much of a lasting, aquatic molluscan fauna. 

In conclusion, the Guadalupe Mountains have stood as a crossroads for 
periodic North American areal molluscan movements. Most representative 
species of these areas still inhabit the Guadalupes, though many forms have 
ceased to inhabit the surrounding areas. Although the molluscan fauna is 
stratified, the majority of the species occur at elevations over 6500 ft in the 
wooded canyons. It should be noted here that most population localities are 
somewhat vaguely referred to in this report for protective purposes. The 
Guadalupe Mountains, due to their unique molluscan fauna and their rela- 
tively small size, offer themselves as a tremendous "laboratory" for solving 
existing molluscan problems. For this reason, as well as others, additional 
care must be exercised to sustain the fragile populations. 

Identified specimens have been catalogued into the Dallas Museum's mol- 
luscan collections and a duplicate set has been deposited with the National 
Park Service at Carlsbad, New Mexico. 

LITERATURE CITED 

Baker, F. C. 1928. The freshwater Mollusca of Wisconsin, Pt. 1, Gastropods. Bull. 

Wis. Geol. Nat. Hist. Surv. 70:326-394. 
Bequaert, J. C. 1966. The molluscan fauna of the Guadalupe Mtns. in Culberson 

Co., Texas. Unpublished, on file with the Carlsbad Caverns National Park 

Headquarters. 
Bequaert, J C,andW. B.Miller. 1973.TheMollusksofthe Arid Southwest with 

an Arizona Checklist. Univ. Arizona Press, Tucson, 271 pp. 
Burch, J. B. 1962. How to Know the Eastern Land Snails. . . . Wm. C Brown Co., 

Iowa, 214 pp. 
Cheatum, E. P., and R. W. Fullington, 1971. The aquatic and land mollusca of 

Texas. Part 1: The Recent and Pleistocene members of the gastropod family 

Polygyridae in Texas. Bull. Dallas Mus. Nat. Hist. 1:1-74. 



110 FULLINGTON 

1973. The aquatic and land Mollusca of Texas. Part 2: The Recent and 

Pleistocene members of the Pupillidae and Urocoptidae (Gastropoda) in Texas. 
Bull. Dallas Mus. Nat. Hist. 1:1—67. 

Cheatum, E. P., R. W. Fullington, and L. Pratt. 1972. Molluscan records from 
West Texas. Sterkiana 46:6-10. 

Fullington, R. W., and W. L. Pratt, Jr. 1974. The aquatic and land Moilusca of 
Texas. Part 3: The Helicinidae, Carychiidae, Achatinidae, Bradybaenidae, 
Bulimulidae, Cionellidae, Haplotrematidae, Helicidae, Oreohelicidae, 
Spiraxidae, Streptaxidae, Strobilopsidae, Thysanophoridae, Valloniidae (Gas- 
tropoda) in Texas. Bull. Dallas Mus. Nat. Hist. 1:1-48. 

King, P. B. 1948. Geology of the Southern Guadalupe Mountains. U.S. Dep. Inter. 
Geol. Surv., Prof. Pap. 215, 183 pp. 

Leonard, A. B., and J. C. Frye. 1962. Pleistocene molluscan faunas and physio- 
graphic history of Pecos Valley in Texas. Rept. Investig. Tex. Bur. Econ. Geol., 
Univ. Texas, 45:1-42. 

Metcalf, A. L. 1970. Field journal of Henry A. Pilsbry pertaining to New Mexico 
and Trans-Pecos, Texas. Sterkiana 39:23-37. 

Metcalf, A. L., and W. E. Johnson. 1971. Gastropods of the Franklin Moun- 
tains, El Paso County, Texas. Southwest. Nat. 16:85-109. 

Pilsbry, H. A. 1939. Land Mollusca of North America (north of Mexico). Acad. 
Nat. Sci. Phila. Monogr. 3 (vol. 1, part 1): 1-573. 

1940. Land Mollusca of North America (north of Mexico). Acad. Nat. Sci. 

Phila. Monogr. 3 (vol. 1, part 2): 575-994. 

1946. Land Mollusca of North America (north of Mexico). Acad. Nat. Sci. 



Phila. Monogr. 3 (vol. 2, part 1): 1-520. 

1948. Land Mollusca of North America (north of Mexico). Acad. Nat. Sci. 



Phila. Monogr. 3 (vol. 2, part 2): 521-1113. 
Pilsbry, H. A.,andE. P. Cheatum. 1951. Land snails from the Guadalupe Range, 

Texas. Nautilus 64:87-90. 
Taylor, D. W. 1966. A remarkable snail fauna from Coahuila, Mexico. Veliger 

9:152-228. 
Walton, M. L. 1963. Length of life in west American land snails. Nautilus 

76:127-131. 



ACKNOWLEDGMENTS 

I first became interested in southwestern mollusks through the late Dr. 
E. P. Cheatum and made my first collecting trip to the Texas Guadalupes 
with him 6 years ago. For more than just teaching me the elements of mala- 
cology, but for instilling in me a love for the Texas mountains, I am grateful. 

Hal P. Kirby, director of the Dallas Museum of Natural History, encour- 
aged our work in the Guadalupes and has supported the molluscan research 
both financially and physically, even to the point of having a special pair of 



MOLLUSCA 1 1 1 

eyeglasses made ". . . to see those little guys with." To Mr. Kirby, the mam- 
malogists, ornithologists, botanists, secretaries, and floor attendants from 
the museum who panted up the mountains with me unselfishly, I am most 
grateful. 

I also want to express sincere thanks to the National Park Service per- 
sonnel: Phillip Van Cleave, Dr. Gary Ahlstrand, Roger Reiche, Norman 
Stephens, and the others who have been most helpful during our stays in the 
mountains. 

Dr. J. C. Bequaert, through the National Park Service, made his unpub- 
lished work on the Guadalupes available and I appreciate this. 

Finally, I wish to thank Mr. Lloyd E. Logan for making his fossil mollusks 
available; Dr. A. L. Metcalf for his personal aid in information and for aid in 
verifying several species; also, to the many malacologists who found strange 
shells on their desks to verify with nothing but thanks for their effort, I am 
most appreciative. 



Plusiotis woodi and Plusiotis 
gloriosa (Scarabaeidae); First Report 
of the Guadalupe Mountains 
National Park, Texas 



RICHARD W. FULLINGTON and DON HARRINGTON, 
Dallas Museum of Natural History, Texas 

Two species of a basically Mexican beetle genus, Plusiotis, were collected 
in the Texas Guadalupe Mountains during 20 to 25 July 1974. Plusiotis 
gloriosa LeConte and Plusiotis woodi Horn were collected in abundance by 
the Dallas Museum of Natural History staff. 

Plusiotis gloriosa was taken from juniper trees in McKittrick Canyon, 
Upper Dog Canyon, and Pine Springs Canyon. Elytra were quite common 
in raccoon scat as were juniper seeds. Most specimens were taken late in the 
afternoon and around lights at night at elevations above 5000 ft (1500 m). 
However, no specimens were found above 6500 ft (1950 m). 

Plusiotis woodi proved to be even more abundant than P. gloriosa. Large 
mating groups were observed in clumps of oak and desert walnut in Pine 
Springs and McKittrick canyons. No specimens were observed above 6000 ft 
(1800 m) elevation and no elytra were found in animal scat. 

The genus Plusiotis is mainly Mexican in distribution, with only 4 of the 
55 known species occurring in the United States. P. gloriosa is known from 
the Davis Mountains, Texas, to Pena Blanca, Arizona, and southward into 
Mexico. P. woodi occurs only in the Davis Mountains, Texas, and at Pinos 
Altos in the Mexican state of Chihuahua (Anon. 1973). Their discovery in 
the Guadalupe Mountains greatly extends the range of these two species 
northward and probably represents the extent of their range in the United 
States. 

The activity of both species is highly moisture dependent (Cazier 195 1). It 
was thought that they were strictly nocturnal until Young ( 195 1) studied the 
habits of P. gloriosa in Arizona. He found that they actively fed during a 24- 
hour period as long as the humidity was high. Rain fell nearly every day of 
our stay in the Guadalupes and the sky was overcast, thus creating excellent 

113 



1 14 FULLINGTON AND HARRINGTON 

conditions for the beetles to emerge from the ground where they burrow 
during dry periods. 

For several reasons, care should be exercised in protecting these two 
species. They are highly prized by professional and amateur collectors and 
even though during short, wet periods they appear very abundant., the popu- 
lations could quickly be depleted. Second, as Young (195 1) postulated, these 
species are relicts from a cooler, wetter era of the Southwest. He felt that 
many more species were present in the United States until the climate in the 
area began drying. Thus, P. gloriosa and P. woodi represent "living fossils" 
in the mountains and merit protection. 

Specimens have been deposited in the Dallas Museum's Entomology Col- 
lection and with the National Park Service, Carlsbad, New Mexico. 

LITERATURE CITED 

Anonymous, 1973. Guide to the U.S. species of Plusiotis. Insect World Digest 
1(2): 14-15. 

Cazier, M. A. 1951. The genera Chrysina and Plusiotis of north central Mexico 
(Coleoptera: Scarabaeidae). Am. Mus. Novit. 1516:1-8. 

Young, F. N. 1951. Notes on the habits of Plusiotis gloriosa LeConte (Scara- 
baeidae). Coleopt. Bull. 11:67-70. 



ACKNOWLEDGMENTS 

We wish to express our gratitude to the museum staff who accompanied us 
to the Guadalupes. Our thanks, too, to the park staff who often had 
"prowlers" around their lighted window screens. 



Notes on the Bionomics and Nest 
Structure of Pogonomyrmex 
maricopa (Hymenoptera: 
Formicidae) 



JAMES V. MOODY and DAVID E. FOSTER, Texas Tech 
University and Museum, Lubbock 

Pogonomyrmex maricopa Wheeler is a widely distributed harvester ant 
common to semi-arid areas of Texas, New Mexico, Arizona, and northern 
Mexico (Cole 1968). Colonies of P. maricopa are characterized by dome- 
shaped mounds surrounded by large bare areas. The species commonly 
occurs in sandy, silt, and loam soils on flat or gently sloping areas. They 
infrequently occupy rocky areas or steep inclines. 

A number of workers have studied harvester ants with the objective of 
control or eradication. The bionomics of a related species, Pogonomyrmex 
occidentalis (Cresson), has been studied (Lavigne 1969; Rogers 1974). The 
only noneconomic works on P. maricopa deal specifically with oxygen con- 
sumption under laboratory conditions (Ettershank and Whitford 1973) and 
home range orientation (Holldobler 1974). 

A study was initiated in 1973 to examine the bionomics of P. maricopa in 
the valley floor habitats of Guadalupe Mountains National Park. The study 
is being conducted in two phases. Phase one, presently underway, is a study 
of nest and colony activities, including nest structure, reproductive cycle, 
foraging behavior, and the effect of abiotic factors on foraging. Phase two 
will focus on the energetic requirements of the species. 

STUDY AREA. The study area chosen for phase one is the floor of West 
Dog Canyon along the northern boundary of the park. The site was chosen 
for the following reasons: its vegetative cover approximates that found on 
most valley floors within the park; it has numerous ant colonies; a deep 
ravine runs through the center of the site that facilitates excavation of 
adjacent colonies; the site is easily accessible. 

METHODS. Observations were made and colonies were excavated during 

115 



1 16 MOODY AND FOSTER 

April, May, June, July, August, and October 1974. Colonies chosen for 
excavation were located within 5 ft (150 cm) of the ravine. They were reached 
by digging laterally from the ravine wall toward the colony. Nest structure 
was mapped and the number of eggs, larvae, pupae, workers, and inquilines 
in each chamber was recorded. 

Foraging activity periods and rates were determined by selecting three 
colonies for observation during their above-ground activity period each day. 
Foraging ants leaving the mound during a 5-minute period were counted at 
hourly intervals. The soil surface temperature during each count period was 
recorded. 

Forage materials gathered were determined by aspirating each returning 
forager into an empty vial; aspirated ants generally responded by dropping 
their booty. The booty of each ant was placed in a separate envelope. One 
hundred returning ants were taken during the July, August, and October 
study periods. Booty was separated into six categories — dry twigs, dirt and 
pebbles, seeds, living plant material, arthropod parts, and bird feces. 

The influence of abiotic factors on above-ground activity was observed 
during all phases of the study. Initial observations suggested that above- 
ground activities of P. maricopa are influenced largely by temperature. 

RESULTS AND DISCUSSION. Pogonomyrmex maricopa constructs a 

dome-shaped mound surrounded by a bare area of 4 to 10 ft (120 to 300 cm) 
in diameter. The mound is constructed from excavated soil supplemented 
with twigs, pebbles, and miscellaneous particles returned by foraging 
workers. Completed mounds stand 12 to 18 in. (30 to 45 cm) high and range 
from 24 to 30 in. (60 to 75 cm) in base diameter. 

Very young colonies lack the characteristic mound and bare area. They 
surround their nest entrance with twigs and debris. As excavation and ex- 
pansion proceed, the mound slowly becomes more typical in appearance. 
Mounds are evident in second year colonies. The development of the bare 
area corresponds with the construction of the mound. 

The nest entrance is located usually 2 to 3 in. (5 to 7.5 cm) above the base of 
the east-facing mound surface. Occasionally, a nest has two separate 
entrances. If so, they are close together and open into the same gallery. The 
entrance is roughly circular and large enough to accommodate two or three 
ants simultaneously. The position of the entrance facilitates warming in the 
early morning and shading in the late afternoon. This allows workers to 
begin foraging early in the morning and to resume foraging earlier in the 
afternoon. 

In 1974 severe drought conditions existed until July, disrupting normal 
mound maintenance. During the drought period, many nest entrances were 
broken open, exposing entire galleries. Such openings never were closed at 
the end of daily above-ground activity. After the advent of rains in July, 
entrances were repaired. Subsequently, all entrances were closed each day 
when foraging was terminated. 



ANT BIONOMICS 



117 



DEP/H 




RELATIVE POSITIONS 
OF VERTICAL SHAFTS AT I" 



Fig. 1. Nest structure and underground distribution of P. maricopa in West Dog 
Canyon, Guadalupe Mountains National Park, 1974. 



The mound and the first 12 to 16 in. (30 to 40 cm) below it are honey- 
combed with dome-ceilinged, flat-floored chambers. Below the honey- 
combed area, two or three main vertical shafts extend downward to a depth 



118 MOODY AND FOSTER 

of 42 to 78 in. (105 to 195 cm) below the mound entrance (Fig. 1). The dis- 
tance between chambers along the first foot of the vertical shafts is usually 
less than 2 in. (5 cm). At lower depths, the chambers are separated by as 
much as 10 to 12 in. (25 to 30 cm). Vertical shafts usually dead end, but some- 
times merge with an adjacent shaft. The largest chambers were found in the 
mound and near the bottom of the nest. 



TABLE 1. Population of Pogonomyrmex maricopa nests excavated in West Dog Canyon, 
Guadalupe Mountains National Park, 1974. 

^fN^t 011 20 April 11 May 9 June 9 July 7 August 19 October 



Eggs 


40 


121 


73 


103 


3 





Larvae 


51 


328 


278 


1331 


470 


46 


Pupae 








83 


1104 


393 





Workers 


13,816 


5932 


1505 


3148 


3442 


2958 


Total 


13,907 


6381 


1939 


5789 


4308 


3004 



Although colonies with approximately equal mound and base area size 
were selected for excavation, populations varied greatly (Table 1). There 
seems to be no correlation between mound size and ant numbers in colonies 
2 years old or older. Growing colonies sometimes abandon nests in favor of 
nearby, larger vacant nests. The age structure of each nest population exca- 
vated was recorded (Table 1). Reproductive activity had begun prior to the 
first excavation as indicated by the presence of eggs and larvae in the nest. 
Pupae were encountered first on 9 July; no teneral adults were present. It is 
interesting to note that egg production peaked during June. The number of 
both larvae and pupae were greatest in July. Winged adults were present in 
July and appeared above ground in large numbers at the time of the first 
rains. Reproduction declined sharply during the latter part of the summer as 
reflected by egg, larval, and pupal populations encountered during August. 
By October, the only immatures present were a few larvae scattered through- 
out the nest. The nest excavated on 19 October 1974 (Fig. 1) was typical both 
in structure and population distribution. 

All arthropods encountered during nest excavation were collected and 
returned to the laboratory. Many were common associates of ants (Wilson 
1971). Inquilines covered in the following discussion are recorded from the 
nest of Pogonomyrmex spp. for the first time. Two species of beetles, 
Echinocoleus n. sp. (Coleoptera: Leptodiridae) and Limulodes n. sp. 
(Coleoptera: Limulodidae), were encountered in the deeper areas of every 
nest. Both species usually were found in galleries packed with eggs and lar- 
vae. Limulodes were found clinging to many of the eggs and larvae. 
Echinocoleus specimens were found running about the gallery floors. 



ANT BIONOMICS 



119 



Abandoned trash and storage galleries in the honeycomb area 
occasionally were inhabited by larvae of Collops sp. (Coleoptera: 
Malachiidae). Collops larvae are predaceous and probably feed on 
scavengers associated with these galleries. Another insect found in the 
honeycomb area was the larva of Gonasida elata (LeConte) (Coleoptera: 
Tenebrionidae), which apparently spends its entire larval life burrowing 
through the mound. 

Foraging activity is affected strongly by surface temperature. The nest 
entrance usually is uncovered and the first workers emerge when the surface 
temperature at the nest opening reaches 21 to 23° C. Activity increases until 
the surface temperature reaches 40 to 45° C and decreases sharply at 
temperatures above 46° C. Activity is reduced greatly at temperatures above 
57° C. 




Fig. 2. Foraging activity and soil temperatures at hourly intervals on a clear day in 
West Dog Canyon, Guadalupe Mountains National Park, 5 July 1974. Dashed line, 
temperature; solid line, numbers of ants. 



On clear days foraging usually was terminated by 1 1 a.m. and began again 
when temperatures dropped to the 48 to 52° C range (Fig. 2). Afternoon 
activity ceased when temperatures dropped to the upper 20s. This bimodal 
pattern of daily activity did not occur on cloudy days when the soil surface 
was exposed only to brief periods of direct sunlight (Fig. 3). 

It was noted on several occasions that rain caused the cessation of foraging 
activity. During light rain, workers stopped leaving the mound; foragers 
returning with booty were unaffected. During heavy rains, foragers returned 



120 



MOODY AND FOSTER 













\ 










280 








/ 
/ 




->». 


\ 






240 








/ 






\ 
\ 




— 










r^.--^-/ 










200 






/ 








\ 




— 


ANTS 














\ 






PER 160 
















KT""\ 


— 


5MIN. 
















X \ 




120 




/ / 












\\ 

\ \ 


— 


80 




/ / 
/ / 












\ \ 


— 


40 




V I I 


I 


I I I I 






I I 


i i N 





50 
45 

40 

TEMP 
35 «c 

30 

25 

20 



12 1 2 

TIME MST 



Fig. 3. Foraging activity and soil temperatures at hourly intervals on a cloudy day in 
West Dog Canyon, Guadalupe Mountains National Park; 6 July 1974. Dashed line, 
temperature; solid line, numbers of ants. 



directly to the nest. Termination of light rain brought a return of normal for- 
aging activity. During the first drought-breaking rains of July, large num- 
itfers of workers massed around their nest entrances, where they imbibed 
water for several minutes at a time. Such behavior did not occur during sub- 
sequent rains in July and August. 

Foraging patterns suggest that populations in the study area forage from 
their mounds in all directions. Ants generally cross the bare area by fairly 
well-defined routes, which lead in four or five directions. Beyond the bare 
area, movement becomes less directional. Foraging distances sometimes 
extend as far as 41 ft (12.3 m) from the nest entrance. However, most forage 
excursions did not extend more than 16 to 20 ft (4.8 to 6 m) from the 
entrance. Foraging is terminated when the first acceptable booty is 
encountered. Returning ants usually follow a more direct route. 

Booty returned by foraging ants during the early afternoons of 9 July, 3 
August, and 18 October were collected and analyzed. On 9 July a few days 
after the rainy season had begun, insect parts comprised 35% of the material 
returned, whereas seeds accounted for 54%. Twigs and pebbles accounted 
for only 8%. On 3 August, seeds and insect parts accounted for only 32% of 
the booty; twigs and pebbles comprised 68%. Mound repair appeared to be 
the prime function of workers visible during the study period. On 18 
October, the colony was engaged in gathering provisions. Sixty-two percent 
of the material returned was seeds, whereas insect parts comprised only 2%. 
Twigs and pebbles accounted for 35%. 

Observations made over the entire season suggest that seeds are gathered 
whenever they are available, and that most plants within the forage range are 
utilized. Insect parts comprise a substantial part of the forage material only 
after the initial summer rains, which trigger a flush of insect emergence. 



ANT BIONOMICS 121 

Twigs, dirt, and pebbles are gathered most frequently during late afternoon. 
These materials are applied directly to mound-building or repair. 

LITERATURE CITED 

Cole, A. C. 1968. Pogonomyrmex Harvester Ants. Univ. Tennessee Press, Knox- 

ville, 222 pp. 
Ettershank, G., and W. G. Whitford, 1973. Oxygen consumption of two species 

of Pogonomyrmex harvester ants (Hymenoptera: Formicidae). Comp. Biochem. 

Physiol 46:605-611. 
Holldobler, B. 1974. Home range orientation and territoriality in harvesting ants. 

Proc. Natl. Acad. Sci. 71:3274-3277. 
Lavigne, R. J. 1969. Bionomics and nest structure of Pogonomyrmex occidentalis 

(Hymenoptera: Formicidae). Ann. Entomol. Soc. Am. 62:1166-1175. 
ROGERS, L. E. 1974. Foraging activity of the western harvester ant in the short grass 

plains ecosystem. Environ. Entomol. 3:420-424. 
Wilson, E. O. 1971. The Insect Societies. The Belknap Press, Cambridge, Mass., 

548 pp. 



Limnology of McKittrick Creek, 
Guadalupe Mountains National Park 



OWEN T. LIND, Baylor University, Waco, Texas 

McKittrick Creek is an unusual aquatic ecosystem. It is a completely 
isolated system surrounded by the Chihuahuan Desert and arid mountains 
of southern New Mexico and western Texas. Such isolation has resulted in 
relatively little human impact until its inclusion into Guadalupe Mountains 
National Park. McKittrick Creek is a significant part of the park which has 
been described as "an island in the sky." Within this canyon, several small, 
relict communities persist. Such isolated systems have much appeal to the 
scientist concerned with community evolution and dynamics. The imminent 
threat to such microcosmic natural laboratories by park development and 
visitor use prompted the aquatic surveys reported here. Financial support 
for this second visit to the canyon was provided by the National Park Service 
preliminary to development of the Master Plan. 

DESCRIPTION OF THE CREEK-CANYON HABITAT 

McKittrick Creek is a small, discontinuous, spring-fed stream with two 
main branches. Situated primarily in the sharp limestone canyons of Guada- 
lupe Mountains National Park, Culberson County, Texas, with a small sec- 
tion in Lincoln National Forest, Eddy County, New Mexico, the principal 
direction of flow is easterly, cutting through the Permian limestone of the 
Guadalupe escarpment where surface flow ends (Fig. 1). This aquatic sys- 
tem provides a unique and isolated resource in the midst of an arid region. 
The creek and its several associated springs and seeps produce a moderation 
of the canyon climate that permits more water-requiring species of ter- 
restrial life to exist. In the protected canyon, woody vegetation includes 
Juniperus, Quercus, Acer, Juglans, Pinus, and Arbutus. Some stream-side 
alligator junipers and ponderosa pines reach diameters of 1 m. 

For the purpose of this study, the McKittrick Creek system is divided into 
three naturally identifiable zones. These are as follows: North McKittrick 
Creek, with surface water flow in both New Mexico and Texas and its south 
fork with the principal spring originating here; South McKittrick Creek, 
from the confluence with North McKittrick Creek to its large headwater 

123 



124 LIND 




- w^ w/^- ■ f '-V^^y) iffS^J rS'y^') >VWi 




Fig. 1. Map of the McKittrick Creek portion of Guadalupe Mountains National 
Park, Texas, and adjacent parts of New Mexico showing location of sampling sites. 



tanks; and McKittrick Creek downstream from the confluence of the north 
and south forks to a point of final subterranean disappearance. A wind-mill 
well is ?t the mouth of the canyon and assumed to be pumping from the 
subsurface creek flow. 



McKITTRICK LIMNOLOGY 125 

Surface flow is usually over a travertine bed of varying thickness, and the 
intermittent creek flow is usually associated with breaks in this bed. Regions 
of surface flow are indicated in Fig. 1. 

I have defined three physically as well as biologically different habitats in 
McKittrick Creek. The first, identified as a run, is a reach of water of up to 15 
cm deep. It has a smooth surface film or, at most, evenly spaced swells. The 
usual bottom substrate in a run is cobblestone from 5 to 10 cm diameter. 
Macro-algae, principally Spirogyra, are often attached. Finely divided 
organic detritus is deposited among the stones. The second habitat type, the 
riffle, has a steeper gradient than the run. The water depth is usually less than 
5 cm. The substrate grades from 15 cm diameter cobblestones at the head to 
small gravel or even sand at the tail. Riffles are quite short, rarely exceeding 2 
m in length. They are typically free of attached macro-algae and organic 
matter deposits. A pool, the third habitat type, does not indicate standing 
water, but defines a wider and deeper portion with slow, usually impercepti- 
ble current. These range up to 5 m in width and 2 m in depth. The substrate 
grades from sand to silt. Large deposits of fine organic detritus are present. 
The bottom frequently has well-developed beds of Chara with marl forma- 
tions. 

The stream-bed gradient is slight in the lower canyon as it is in North 
McKittrick Creek up to the south fork, and in South McKittrick Creek up to 
the abandoned Grisham-Hunter Lodge (Fig. 1). The watercourse of North 
McKittrick Creek (Stations A and B, Fig. 1) below the south fork is the most 
exposed in a broad, open canyon. The south fork of North McKittrick Creek 
is well shaded by woody and herbaceous vegetation with maidenhair ferns 
(Adiantium capillus) providing much cover. Protective stream-side vegeta- 
tion of the main canyon and of the portion of South McKittrick Creek below 
the Grisham-Hunter Lodge is dominated by saw-grass (Cladium 
jamaicense) as well as deciduous trees. Above this point, South McKittrick 
Creek has less protection by woody vegetation but several reaches flow 
among large boulders and are shaded with maidenhair fern. 

METHODS 

Data and observations contained herein are based on a series of visits. The 
first was November 1967, followed by visits in May 1969, April 1971, and 
November 1972. 

Collection stations were selected to provide a sampling of each major 
habitat type as well as stream portions having real or potential human 
impact. 

Most chemical analyses were done in the field using the Engineer's 
Laboratory, Hach Chemical Co. Specific conductance and total residue 
were determined according to Standard Methods for the Examination of 
Water and Waste Water (Am. Public Health Assoc. 1965, 1971). Iron, 
copper, lead, and zinc contents were determined by atomic absorption and 
total organic carbon, by combustion and infrared analysis. 



126 



LIND 



jb 2 
U 3 



q 














CN 


CN 




CO 


CO 


oo 




o 
"3- 


o 
-1- 


o 

to 

m 


o 

OS 
CN 


CO 

CN 


CO 


OS 


o 
<* 

CO 


NO 


NO 


lO 


io q 












CN 






CO 


00 


r^ 


r*» r^ 


o 

CO 
co 


to 
cs 
CO 


o 
m 
co 


o 

co 




r-^ 




O 
cn 


CO 


vd 



q q 

^ Tf CO CO 



q 
o t> 











CN 






CN 


CO 


o o 

CO CO 
CN CN 


o 

CN 


O 
CN 


o 


CO 


Tt 


o 


Os 


CO 


OS so 
CN CN 


O 

o 

co 


O 
to 


Os 




CO 
CN 








O 




o 






d 




m 





lO 






r-^ 


o 
m 

CN 


O 
CN 







CO 


«o 












lO 


q 




CN 


ft 


rN 




CO 

q 


r-^ 


CN 
CN 


O 

cs 


CN 

r- 


O 
r- 
CN 


CN 


so 
q 


Os 


o 
o 

CO 


OS 

CO 


NO 

<* 


O 

CN 




CO 

r-- 


CO 


CN 
sO 
CN 




o 
o 

CO 






CO 

q 


vq 


o 

CO 


CO 
NO 


NO 
CN 


CO 


«* 




SO 


NO 
CN 


o 
to 
cn 


O 
O 
CO 


o 
r- 

CN 


CN 

CO 


q 


VO 


o 

OS 


1— 1 


CO 


o 
cn 




CO 
C-; 




CN 

r- 

CN 




CN 

i— < 

co 




CO 


no 




o 

CO 


CN 

q 


no 

CO 


NO 
CN 




CO 




SO 

tO 




CN 
P- 




CN 






o 


CN 


SO 



CN OS 

r- so 



OS — < H O) H 

so r- r- r- r- 



IO ^t H IO Tf H 



3 



>o Tt »o lO 



>» 


0) 




8 




T3 




03 


'3 


53 
o 




o * 


■a &c 
73 « 


3 ^2 CO -^ CN 

cx o _, O hs 
c/d Q ® ivo oo 


H 


H 


u o 



McKITTRICK LIMNOLOGY 



127 



o 

O oo 



co ^h 



r- <n o 

o o o 



Q c3 



4) 3 fl 



gfS 



128 LIND 

Quantitative benthos samples were by Surber sampler with 3 square ft 
sampled at each station. Qualitative samples were by sweep net and hand- 
picking. Fishes were collected by habitat seine. Plankton from the rock 
tanks were collected with a 20-mesh Wisconsin net. Identification of most 
of the vascular aquatic plants and aquatic-associated plants was confirmed 
by Helen and Donovan Correll. Mosses were identified by L. J. Gier. 

RESULTS AND DISCUSSION 
Physical and Chemical Properties 

Because of the intermittent nature of McKittrick Creek, water tempera- 
tures do not have a continuous, gradual rise downstream. The spring source 
of North McKittrick Creek's south fork (station C) was cool at 15 and 16° C. 
In the open, shallow, travertine-lined tanks downstream (station B), this rose 
to 20 and 25° C, then dropped, after further subterranean flow, to 14° C at the 
last downstream station (A) (Table 1). This same irregular pattern of 
temperature was followed in South McKittrick Creek. 

The water chemistry reflects the limestone substrate of the region. This is a 
well-buffered calcium carbonate-magnesium carbonate system with a 
greater calcium to magnesium ratio in the water of South McKittrick Creek 
than in North McKittrick Creek. Bicarbonate is the principal anion in both 
creeks. In the more exposed waters, high rates of photosynthesis deplete the 
carbon dioxide-bicarbonate system with a consequent rise in pH and the 
production of the insoluble carbonate ion resulting in heavy marl forma- 
tion. Sulfate, the other anion present in significant quantities, is slightly 
more abundant in South McKittrick Creek water than in North McKittrick 
Creek water. 

Dissolved oxygen concentrations at all sampling stations, except the 
uppermost tank at station 8 of South McKittrick Creek, were always con- 
siderably greater than those shown to cause stress on any forms of aquatic 
life. This is to be expected in the relatively cool, shallow, and flowing water. 
However, when taken as a percentage of saturation, South McKittrick 
Creek waters always were less saturated than North McKittrick Creek 
waters, suggesting a greater organic matter load in the former. No diel oxy- 
gen studies (the reported data being for mid-day) were conducted. 
Consequently, I do not know if oxygen concentrations are significantly 
lowered at night. However, the biota present demonstrate that any decrease 
is not to a critical concentration. 

The uppermost rock tank at station 8 of South McKittrick Creek had a 
mid-afternoon oxygen concentration of 0.2 mg l 1 . Because of its location 
and morphometry configuration, this station collected large quantities of 
organic matter. The orientation of the vertical canyon walls coming directly 
out of the water provides almost constant shading and protection from the 
wind. Other physical and chemical conditions associated with this sheltered, 
organir-rich system were the lower temperature, low pH, and higher free 
carbon dioxide. 



McKITTRICK LIMNOLOGY 129 

Nitrogen, phosphorus, and occasionally carbon and silica are of concern 
to limnological investigations because of their confirmed role as elements 
often, or occasionally, limiting the quantity of biological production. In 
well-buffered carbonate systems such as these, carbon is probably always 
superabundant. Dissolved silica also was present in relatively high quanti- 
ties, although there is little information available on limiting effects of silica 
concentrations especially in streams. These concentrations, 6.4 to 8.8 mg l 1 , 
are similar to those shown to be limiting in a central Texas reservoir (Kimmel 
and Lind 1972). However, they are approximately four times greater than 
the ambient concentrations of Douglas Lake, Michigan, which I found to 
cause no limitation (Lind, unpublished). The dissolved silica concentration 
of any water is inversely related to the population size of active diatoms. 
Though not sampled, diatom populations in both forks of McKittrick Creek 
were large, as evidenced by the color and texture of the rock and travertine 
substrate. It is probable that a large portion of the silica present in these 
samples was tied up in the silicon tests of active diatoms. The high concen- 
tration of dissolved silica at station B probably indicated the decline of a 
large diatom population in the travertine runs immediately upstream. 

Nitrate-nitrogen concentrations increased downstream in South 
McKittrick Creek from undetectable at the upper rock tank to 0.35 mg l -1 
at station 3 below the Grisham-Hunter Lodge. The threefold increase at 
station 3 over station 5 above the lodge causes one to suspect human sources, 
possibly septic seepage associated with the lodge. The high concentration of 
nitrate-nitrogen in the spring water of North McKittrick Creek is not 
unusual for many springs of the region. Three nearby springs on the east face 
of the Guadalupe Mountains were analyzed during the 1971 study. These, 
Manzanita,Choza, and Upper Pine springs, had nitrate-nitrogen con- 
centrations of 20, 50, and 50 mg H, respectively. These concentrations 
approach or exceed the recommended drinking water limit of 45 mg l _l 
established by the American Public Health Association (1971). Smith 
Spring, which presently is diverted into Manzanita Spring, had a much 
lower nitrate-nitrogen content of 0.46 mg H. 

Phosphate-phosphorus concentrations were measured at stations 3 and 5 
to gather further data on possible human impact. The results were similar to 
those of nitrate-nitrogen. The downstream station had a phosphorus con- 
centration 5 times greater than the upstream. 

Little is known about normal concentrations of heavy metals in non- 
poluted waters. Riley (1939, cited by Hutchinson 1957) found the range 
of copper of Connecticut lakes to be from 0.009 to 0.215 mg l 1 . Atkins 
(1933, cited by Hutchinson 1957) found 0.0 to 0. 036 mgl 1 in several English 
rivers. A small central Texas reservoir had a mean concentration of 0.06 mg 
L 1 (Lind 1974). Several springs in Big Bend National Park, Texas, had 
copper concentrations of 0.03 to 0.04 mg l 1 and Tornillo Creek had a range 
from 0.02 to 0.03 mg l 1 (Lind and Bane, unpublished). It is probable that the 
value of 0.02 mg l 1 for all sampled stations in the McKittrick Creek system 



130 



LIND 



is not exceptional. Hutchinson (1957) cited several authors regarding the 
toxic copper concentrations for a variety of aquatic life. The McKittrick 
Creek values are all less than any reported toxic concentration. 

Hutchinson (1957) suggested that zinc, though little studied, is probably 
present in quantities equal to or greater than copper. The zinc concentration 
of McKittrick Creek waters is less than copper and of the same order of 
magnitude and in the same copper: zinc ratio as found by Morita (1955, cited 
by Hutchinson 1957) for high Japanese lakes. The small central Texas 
reservoir had a mean zinc concentration of 0.1 mg l -1 (Lind 1974) or 
approximately double the copper concentration. Concentrations inTornillo 
Creek were 0.01 to 0.02 mg 1 _1 , one order of magnitude greater than in 
McKittrick Creek. It appears from this that ambient zinc concentrations in 
McKittrick Creek are exceptionally low. The concentration of lead was 
below the detection limits of our instrumentation (1 mg l -1 ). This is con- 
sistent with very limited data available on this element. The total iron 
values, 0.02 mg I' 1 , are, as with zinc, approximately one order of magni- 
tude low when compared to data for Swedish lakes (0.09 to 0.160 mg 1 -') 
(Rodhe 1948), for Lindiey Pond, Connecticut (0.170 mg l' 1 ) (Hutchinson 
1957), for a small central Texas reservoir (0.30 mg 1~') (Lind 1974). and for 
Tornillo Creek (0.1 to 0.2 mg 1 ' )(Lind and Bane, unpublished). 

The Biota 

The aquatic and streamside flora of the creek, is varied, most are 
perennials. The transition from aquatic-associated plants of the narrow can- 
yon floor to the xerophytic forms of the canyon sides is abrupt. The criterion 
for inclusion as an aquatic associated plant in Table 2 was the listing of that 
taxon in Aquatic and Wetland Plants of Southwestern United States 
(Correll and Correll 1972). 



TABLE 2. Stream-side and aquatic plants of McKittrick Canyon, Guadalupe Mountains 
National Park. 



Equisetum laevigatum (Horsetail) 
Adiantum capillus (Maidenhair fern) 
Bryum turbinatum (Turbin moss) 
Hygroamblystegium irriguum 

(Spring moss) 
Potomogeton illinoensis (Pond weed) 
Eleocharis montevidensis (Spike rush) 
Cladium jamaicense (Sawgrass) 
Carex microdenta (Sedge) 
Carex hystericina (Porcupine sedge) 
Juncus interior (Rush) 
Juncus dudleyi (Rush) 



Agrostis semiverticillata (Bentgrass) 

Leersia sp. (Cutgrass) 

Glyceria striata (Fowl Manna-Grass) 

Rorippa naturtium-aquaticum (Water Cress) 

Aquilegia chrysantha (Columbine) 

Galium microphyllwn (Bedstraw) 

Valeriana texana (Valerian) 

Senecio sp. (Groundsel) 

Najas sp. (Water nymph) 

Spirogyra sp. 

Chara sp. (Stonewort) 

Nitella sp. (Stonewort) 



McKITTRICK LIMNOLOGY 



13 



3-3 



"3 £" 

C OS 

IS 



3 C 

O o 

^ t 

CJ o3 

•£3 <u 

•"£ > 



1 s* 

2 o 



3 T3 

Ox 



pq S 

< 



>^ 


03 




<L> 


00 




43 


— 




6 


"o 




3 


03 
0) 










cu 


c 
o 




r-3 


T3 




o3 


C 










o3 






T3 


c2 


0) 


D 


<D 


4> 


> 


»H 


"•+3 


d> 


U 


^ 


£ 


j* 


'J 


03 


# o 


c 


X 




o3 

3 


cd 




o 


ai 


5 



X 



X 



X X 



.ts «> .2 

% w s 



o 

J= -c ^ 

■*■* l - ^ aj 

S • - o 

z ^ u 



























« 




2 
























"3 


-=; 




"^ 
























^> 


Cl 




Q 


s/3 




Q 
























* S- 


1 




| 


03 




Kl 












<3 

-2 


•5 




Qj .$3 


3 | 

o 


3 


■9 

bo 


a, # 
o S? 

6 £ 

2 ^ 

s ^ 


a3 


o 



Jp 

O 


O 
3 
O 


a> 

a 

>, 
Ph 


.-3 
O 

u 

3 


03 

o 

CJ 

aj 

M 

o 


03 

-a 
o 
ex 

E 

< 


-5 


03 ?\ 

o 
U 


U 

o 

Ta- 
rt 


5, -5 


u 


H 




Z 


< 




§ 




u 








c 





132 



LIND 



"8 



o> 


5 


2 


W 


U 




M 




<J 




•c 






C 


2 


o 

'■2 * 



a <N 

S o o 

* u 3 






X X X X 



X X 



■S3 3 

s -c 

k3 O 



< > 



McKITTRICK LIMNOLOGY 133 

With a few exceptions, the fauna is typical of more northerly, spring-fed 
streams. This supports the hypothesis of a relict biota, left in isolation by the 
northerly retreat of montane climates and associated biotic communities. 
The three species of fish (Table 3) were apparently imported by the earlier 
landowners for sport fishing. The yellow-belly sunfish (Lepomis auritus) is 
at the limit of its known natural range* The green sunfish (L. cyanellus) was 
probably a contaminent of this sunfish stocking. The rainbow trout (Salmo 
gairdneri) has been successful in the creek although not reaching large sizes. 
Populations of trout were found at the lower downstream stations where the 
water temperatures exceed those in which trout are capable of effective 
reproduction. It is probable that these represent a downstream wash at 
periods of high water, and the spawning must be restricted to the colder 
upstream portions. 

The rock tanks at the head of South McKittrick Creek with cold, some 
with low oxygen waters, though depauperate in benthos, had an interesting 
plankton fauna. Quant ative samples were not taken but dense populations 
of three species were present. Ectocy clops phaleratus stores large quantities 
of orange-yellow oil droplets. This is in the same manner as arctic copepods 
as well as those observed by me in alpine lakes of the Rocky Mountains. This 
is not to be confused with a pink carapace pigmentation also observed in 
some high-altitude species. This oil pigmentation, as well as the dense 
population, contributes an orange hue to the water which may at first be 
attributed mistakenly to organic matter stain. It would be extremely 
interesting to determine the source of inoculation of these plankton into this 
uppermost tank. 

Several detritus-associated taxa are common in the tanks and pools 
holding leaf litter. Ostracods are abundant, especially at station 4 where the 
population exceeded 2300 organisms nr 2 (Table 3). This is also the station 
that experiences the suspected nutrient enrichment. Station 4 also had the 
maximum planaria population density. The amphipod, Hyalellaazteca, was 
present throughout except at the lowermost station 5. The horsehair worm, 
Gordius sp., is restricted to organic rich pools of middle to upper South 
McKittrick Creek and has been found in small numbers on each of the 
sampling dates. 

The insect fauna (Table 4) collected consisted primarily of larval or pupal 
stages. With a few exceptions, no effort to identify beyond family was made. 
Seven families of Diptera were found in South McKittrick Creek and five, in 
the smaller North McKittrick Creek. At least five recognizable genera of 
Tendipedidae (Chironomidae) were found. Simulium sp. populations in the 
riffles were large, exceeding 18,000 organisms nr 2 at station 3.Tendipedids 
were found throughout the system, with one genus being found in sub- 
stantial numbers at all quantitative stations. 

Trichoptera were present in all but the swiftest water throughout. Two, 
Notiomxia sp. and Genus A (Ross), were found only in North McKittrick 



134 



LIND 



0) 








93 




s 




0> 








cd 




3 




& 




,_ 




V 




EX 




C/5 


oj 






Oj 


ed 


Xj 


-a 


5 


a; 


3 


a 


c 


E 


cd 


S! 


71 


,c 




o 


«1 


cd 


T3 


u 


<u 


c 


> 


o 


C-J 


■d 




■- 


c 


3 


cd 


o 


3 


'*- 


o 


<u 




H 




9i 


M 


£ 


7?. 


Cd 

X 




ed 


cd 




Cj 







Ts 



•Z c 



■X! 


■ 




O 


3 


'O 


2 


CX> 


o 




s 


£ 


a> 


o 


a, 
3 




rj 


cd 


-a 


73 


ed 


-a 


3 




a 


a> 




(fl 


^ 


O 


u 


a, 


Oj 
^ 


E 


u 


o 


-^ 


U 


•J 


!/5 


■rj 


in 


£j 


d 


5 


£ 

rrl 



(5 


e 


3 


cd 


ed 


3 




cr 




C 


o 





o 


s 


wj 




c 


c 








o> 





o 

c 


a> 


o 


cj 






£ 



s ° 
ol 



" T3 

W .S 

«x 



o ^ 

5 H o 

^ U 42 






o 


^o 


■^- 00 


C^ 


en 


r- 


(N 


lO OO 


n 


<tf 


<N 


(N 


vo 


r-H 





WO 


r|- <N co 


ro 


o 


•O C-- Tt 


(N 


o 


*H 





lO >0 M 

o t-~ <n 



<N CM 

m t— i 



X X 



X 



X 



cd a> 

3 '53 



5 

3 s 

2 £ 



V3 Cd 



< fflU Q W 

t/5 CO 0O C/3 C/3 

3 3 3 3 3 

3 C C C C 



if 

o 



<U cd iz 

H H on 



3 TJ oc 

■S S .2 

g. o N 

p Wl N 

e* ° ^ 



X X 



.Q cd 



u 



o 



McKITTRICK LIMNOLOGY 



135 



« "8 

-5 § 

"a & 



O 

'■a «r> 



X 



X 



B * £ 

* a 9 



r- oo 

<N r-l 



CO o, 
CM <N 

CO -H 






X 



X 



X X 



c 


<u 


fc 




-S 










§ 


id 


«« 
o 


<u 


-, 


a> 


d 






g 


o 


ex 



03 

"2 
a. 


1 




5 


& 

& 
£ 


c3 

."2 

03 
U 

o 

E 




o 


<fc 


o 


<^ 





Sc 


5' 






&3 




SC3 


■a 


^ 




ffl 




3 




x 






-J 




136 



LIND 



73 Br 



2 



Us 



s 



vs 



§ § O 3 o 

2|8-i1S^|s 

Q en 5 J5 3 <£ 

^ Oh QJ W U O 

o 

u 



?S 



<3 



McKITTRICK LIMNOLOGY 137 

Creek. Heticopsyche sp., Hydropsyche sp., Agraylea sp., and Athripsodes 
sp. were present only in South McKittrick Creek. Damselflies (Zygoptera) 
and dragonflies ( Anisoptera) were found in both systems. Archilestes sp. was 
found only in North McKittrick Creek, whereas Argia sp. was common in 
both systems. The seclusion of dragonfly larva in bankside vegetation and 
the method of sampling could have led to the low population estimates of the 
libellulids. Baetid mayflies (Ephemeroptera) were in all stream zones, with 
Choroterpes sp. present at all quantitative stations. Larval and adult 
beetles were present throughout in the qualitative samples as was the 
hemipterian, Belostoma sp. Subsequent to my last collection, Harley P. 
Brown (pers. comm.) has collected in the lower canyon and has identified 
three dryopid beetles (Helichus confluentus, H. saturalis, and H. 
triangularis), four elmids (Elsianus sp., Heterelmus obesa, Micro cylloepus 
sp., and Neoelmis sp.), one limnichid {Lutrochus luteus), and represen- 
tatives of Gerridae, Naucoridae, and Veliidae. 

The occurrence of several southerly range extensions of insects supports 
the concept of a relict fauna. Psephenus sp. is normally found to the north- 
east of the McKittrick region. The odontocerid caddisfly, known only as 
Genus A by Ross, has been described only from northern California, and the 
georyssid colopteran, Georyssus sp., is reported from the northern Rocky 
Mountains. 



TABLE 5. Mean, maximum, and minimum species diversity of six quantitative sampling 
stations, 1969 and 1971, McKittrick Creek, Guadalupe Mountains National Park. 



Sampling 




1969 






1971 




Station 


Mean 


Maximum 


Minimum 


Mean 


Maximum 


Minimum 


A 


- 


- 


- 


2.5 


3.2 


0.2 


C 


- 


- 


- 


2.5 


4.0 


0.2 


2 


2.9 


3.8 


0.5 


- 


- 


- 


3 (riffle) 


1.7 


3.9 


0.1 


2.9 


3.6 


0.2 


4 


3.0 


4.3 


0.3 


- 


- 


- 


5 


2.6 


4.1 


0.2 


3.0 


3.6 


0.2 


7 


- 


- 


- 


2.0 


3.1 


0.1 



Species diversity indices reflect community stability. A low index sug- 
gests either a youthful, very small, or a disturbed system. Disturbances, 
natural or manmade, tend to bring about a decrease in the index. Most 
diversity data for McKittrick Creek (Table 5) are rather high, especially con- 
sidering the small size of the creek habitats. Because benthos are relatively 
permanent (until their life cycle is complete) inhabitants of stream systems, 
they are continuous monitors of any stresses placed on the system. Cnanges 
in their diversity provide a means to detect stresses that may be missed by 
periodic sampling of biological, chemical, or physical parameters. 



138 LIND 

Stations 3 and 5 were sampled for diversity in both 1969 and 1971. There 
was a slight (though probably insignificant, cf. maximum and minimum) 
increase in 197 1. The increase at station 3 in 1971 may be attributed to the 
fact that this riffle is immediately below a ford crossing used by occasional 
jeep traffic up to about 1970. It is advisable to continue monitoring for 
changes in diversity as an expression of environmental degradation. 

The standing crop of stream organisms is not usable for evaluation of 
water quality because large populations of more tolerant organisms may 
exist even under stress. However, it does provide insight to the quantity of 
organic nutrients available. These may be available as either allochthonous 
detritus or autochthonous production. Generally, small aquatic ecosystems 
are more dependent upon the former. This is probably true of McKittrick 
Creek as the canyon topography provides "funneling" of the downhill-mov- 
ing plant production to the stream. However, the large beds of stream-side 
Cladium jamaicense, Chara sp., and Spirogyra sp. in the pools and runs, as 
well as the large diatom populations, each contribute significantly to the 
food base. 

No biomass data on benthos were gathered. Numeric standing crop data 
were high. The mean for all South McKittrick Creek stations is approxi- 
mately 7000 organisms/ nr 2 , with a maximum for station 3 of 20,000 
organisms/ nr 2 . These numbers are high when compared with data given by 
Macan (1963) who reported mean numeric standing crops of English 
streams to be from 3000 to 5000 organisms/ nr 2 . 

The results of these studies confirm that the McKittrick Creek ecosystem 
is a unique and valuable resource. The opportunity for scientific discovery 
here is great. Sound, carefully regulated research activities should be 
encouraged as there is yet much to know about the structure and dynamics of 
this stream. 

LITERATURE CITED 

American Public Health Association. 19b:>. Standard Methods for the Ex- 
amination of Water and Wastewater. Including Bottom Sediment and Sludges. 
American Public Health Association, Inc., New York, 12th ed., 769 pp. 

1971. Standard Methods for the Examination of Water and Wastewater. 

American Public Health Association, Inc., New York, 13th ed., 874 pp. 

Atkins, W. R. G. 1933. The rapid estimation of copper content of sea water. J. 
Mar. Biol. Assn. U.K. 19:63-66. 

Correll, D., and H. Correll. 1972. Aquatic and Wetland Plants of the 
Southwestern United States. Environmental Protection Agency, Washington, 
D.C., 1777 pp. 

Hutchinson, G. E. 1957. A Treatise on Limnology. John Wiley & Sons, Inc., New 
York 1:1-1015. 

Kimmel, B. L., and O. T. Lind. 1972. Factors affecting phytoplankton produc- 
tion in a eutrophic reservoir. Arch. Hydrobiol. 71:124-141. 
jnd, O. T. 1974. The Relationship of Electric Power Station Thermal Circulation 



McKITTRICK LIMNOLOGY 139 

to the Biological Productivity of Reservoirs. Project completion report, Project 

No. B-091-TEX, Office Water Resources Research, 1 10 pp. 
Macan, T. T. 1963. Freshwater Ecology. John Wiley & Sons, Inc., New York, 338 

pp. 
Morita, Y. 1955. Distribution of copper and zinc in various phases of the earth's 

materials. /. Earth Sci. Nagoya Univ. 3:33-57. 
Riley, G. A. 1939. Limnological studies in Connecticut. Part II. The copper cycle. 

Ecol. Monogr. 9:66-94. 
Rodhe, W. 1948. Environmental requirements of freshwater plankton algae. 

Symb. Bot. Ups. 10(1):1-149. 



The Quaternary Vertebrate Fauna of 
Upper Sloth Cave, Guadalupe 
Mountains National Park, Texas 



LLOYD E. LOGAN and CRAIG C. BLACK, The Museum, 
Texas Tech University, Lubbock 

Upper Sioth Cave is located in the extreme northwestern corner of 
Culberson County, Texas, at an elevation of 2000 m and approximately 2.5. 
mi. northwest of Guadalupe Peak. Vertebrate material obtained from 
excavations of Upper Sloth Cave provides an excellent opportunity to study 
the vertebrate faunal evolution of the southern Guadalupe Mountains from 
approximately 13,000 years Before Present (BP) to the present. A previous 
excavation of Upper Sloth Cave, then called High Cave (Mera 1938), 
yielded few vertebrate remains, but some perishable archaeological 
material was found. Previous work in the southern Guadalupe Mountains 
has shown that the Sangamonian-Wisconsian faunas contained extinct 
genera such as Nothrotherium (Ayer 1936; Van Devender et al. 1977a) and 
extant genera such as Marmota (Stearns 1942; Schultz and Howard 1935) 
and Sorex (Harris 1970b; Logan 1975) that are found only farther north, at 
higher elevations, or in more mesic habitats than now exist in the southern 
Guadalupe Mountains. 

METHOD OF STUDY 

The bones of the vertebrates were identified to the lowest possible 
taxonomic level. Specific identifications of most mandibles and maxilla with 
teeth were possible. Because of the extremely fragmentary nature of most of 
the postcranial material, identifications were not attempted except on the 
Serpentes. 

The habitat preferences and environmental interpretations are based on 
modern literature reports regarding the species represented in the cave 
deposits. 

The vertebrate specimens are cataloged in the vertebrate paleontological 
collections of The Museum of Texas Tech University (TTU-P), Lubbock, 

141 



142 



LOGAN AND BLACK 



under the locality number TTU-TEX-2. The molluscan fauna (Table 1) is 
deposited at the Dallas Museum of Natural History. 



TABLE 1. Mollusca from Upper Sloth Cave. 



Taxa 




Depth 


in cm 






0-10 


10-20 


20-30 


30-40 


Discus cronkhitei 


X 








Gastrocopta ashmuni 




X 






G. pellucida parvidens 




X 






Glyphyalinia indentata paucilrata 


X 


X 


X 




Helicodiscus eigenmanni 


X 








H. s. singleyanus 




X 


X 


X 


Holospira pityis 


X 




X 




Holospira sp. (immature) 


X 


X 






Metastoma roemeri roemeri 


X 


X 




X 


Oreohelix socorroensis socorreonsis 




X 






Oreohelix sp. 


X 


X 






Pupilla blandii 




X 






Pupilla sp. (immature) 


X 


X 


X 




Rabdotus sp. (immature) 


X 








Succinea sp. (immature) 




X 






Vallonia sp. (immature) 




X 







PRESERVATION OF BONE 

The bones of Upper Sloth Cave are well preserved with little or no 
mineralization, but with a high degree of breakage. The fragmentary nature 
of the bones and the fact that very little material present is larger than a jack- 
rabbit suggest that the major bone accumulations were by small mam- 
malian carnivores and predatory birds, although some of the mammals, such 
as Neotoma and Bassariscus, certainly live in and around the cave today. 



STRATIGRAPHY 

Two trenches were excavated in the front chamber of Upper Sloth Cave 
(Fig. 1) during the summer of 1974. Due to the very different nature of the 
deposits, the only correlation between the two trenches is based upon the 
appearance of Sorex cinereus and Cryptotis parva in the 30 to 40 cm level of 
trench 1 and in the 10 to 20 cm level of trench 2 (Table 2). The 30 to 40 cm 
level of trench 1 has been radiocarbon dated at 1 1,760+610 BP (A- 15 19), on 
artiodactyl fecal pellets from an adjacent, but previously excavated, trench 
(Van Devender et al. 1977a). 

Trench 1, located near the west wall, shows the following stratigraphic 
units. Unit 1 is from the surface to a depth of 15 cm where it makes a 
hummocky and somewhat blended contact with unit 2. Unit 1 is composed 



VERTEBRATE FOSSILS 



143 




& 





-oitti 


> 

(0 


£ | 


u 


2 S o 




i 3 

Ol x 


£ 


Ceilim 
Exist! 

1971 1 
Slope 

Bould 







5i 


>J O ? Ol 

• X _j o 


k 


« 9 "• 2 ■" 


0) 


rvey 
leve 

oyd 


a 


3 (fl Q _l Z j 
(A Q 


a 




3 



■s 

p 

a 



9 
> 

u 
o 
% 

P 

a 

> 

u 

o 

CO 
t-l 

1) 
O- 
Q, 
P 



144 



LOGAN AND BLACK 



TABLE 2. Mammals from Upper Sloth Cave. 



Taxa 








Depth in cm 








Trench 1 




Trench 2 






0-10 


10-20 


20-30 


30-40 


Sorex cinereus 


30-40 




X 






Cryptotis parva 


10- 


-20 




X 








30-40 










Notiosorex crawfordi 






X 


X 


X 


X 


Myotis velifer 






X 








My otis sp. 






X 








Lasionycteris noctivagans 








X 






Eptesicus fuscus 








X 


X 




Plecotus townsendi 






X 








Antrozous pallidus 






X 




X 




Nothrotherium shastense 


25- 


-45 










Sylvilagus sp. 






X 


X 






Lepus sp. 






X 








Marmota flaviventris 








X 






Spermophilus variegatus 






X 


X 






Thomomys bottae 






X 


X 






Pappogeomys castanops 






X 








Peromyscus spp. 






X 


X 


X 


X 


Peromyscus eremicus 






X 


X 






Neotoma cinerea 






X 


X 


X 




Neotoma mexicana 






X 


X 


X 


X 


Neotoma albigula 








X 






Neotoma micropus 








X 






Microtus mexicanus 






X 


X 


X 


X 


Bassariscus astutus 






X 


X 






Mustela frenata 






Surface of i 


i spoil heap near trench 2 




Felis concolor 






X 









of many cobble and gravel-sized limestone fragments, much plant material, 
limited archaeological material in the form of quids and charcoal frag- 
ments, and much fine grey dust. Small artiodactyl fecal pellets are present, 
but other vertebrate material is sparse. Unit 2 differs from unit 1 in that the 
archaeological material and most plant material are absent. The dust is more 
of a reddish tan, but not the typically red clays of most cave deposits. Unit 2 
lies from 15 to 25 cm below the modern surface of the cave and makes a 
rather sharp but hummocky contact with unit 3. Unit 3 lies from 25 to 45 cm 
below the surface and is composed primarily of dung balls of Nothrotheriops 
shastense (Shasta ground sloth) in various stages of decomposition (Van 
Devender et al. 1977a). The majority of identifiable vertebrate remains 
found in this trench are from stratigraphic unit 3 at the 30 to 40 cm level and 
consist primarily of mandibles of Sorex cinereus and Cryptotis parva. Also 
present in unit 3, and increasing in mass toward the center of the chamber, 



VERTEBRATE FOSSILS 145 

are sticks and small logs of Pinus edulis (limber pine). Unit 3a is an ap- 
parently wind-accumulated layer of leaf litter at the 40 to 45 cm level. This 
unit is present only in a local area in the east end of the trench directly under 
the largest concentration of small pine logs. One partial skull and associated 
mandibles of Sorex cinereus were recovered from this unit. Unit 4 consists of 
limestone cobbles averaging 20 cm in diameter, with the reddish dust tilling 
the cracks between the rocks. Little vertebrate material was recovered from 
the 45 to 70 cm levels. Trench 1 was terminated at this level. 

Trench 2, located at the mouth of a small passage on the east side of the 
main chamber (Fig. 1), contains the following stratigraphic units. Unit 1, 
from the surface to 6 or 7 cm, is composed of a dry, grey dust mixed with 
many pieces of broken flowstone. Fragmentary vertebrate material is 
extremely abundant. Unit 2 is a calcite cement layer up to 3 cm thick along 
the walls of the passage, but interrupted toward the center of the passage. 
Some vertebrate material is incorporated within the cemented matrix. Unit 3 
is from 6 to 10 cm below the surface of the trench to a depth of approxi- 
mately 40 cm, and is composed of a typically reddish and slightly damp clay 
fill with many gravel-sized limestone fragments. Vertebrate material is 
extremely common throughout the trench, but nearly all of it is of a frag- 
mentary nature. 



SPECIES ACCOUNTS 

Sceloporus poinsetti Baird and Girard, Crevice Spiny Lizard 

Material.— Right dentary (TTU-P-8312); right maxilla (TTU-P-8313). 
Discussion. — S. poinsetti is a relatively common lizard in the vicinity of Upper Sloth Cave and 
is found throughout the rocky, arid areas of the southern Guadalupe Mountains. S. poinsetti 
does not indicate any change from present environmental conditions. 

Sceloporus undulatus (Latreille), Eastern Fence Lizard 

Material. — Two right dentaries (TTU-P-8315-8316); seven left dentaries 
(TTU-P-83 17-8323); four left maxilla (TTU-P-8324-8327); five right maxilla 
(TTU-P-8328-8332). 

Discussion. — S. undulatus has previously been reported only twice from prehistoric localities in 
the southwestern United States (Holman 1970; Gehlbach and Holman 1974). S. undulatus is an 
abundant lizard in the vicinity of Upper Sloth Cave and occurs in all biomes of the southern 
Guadalupe Mountains. This species is not useful as an ecological indicator because of its 
ecological plasticity. 

Urosaurus ornatus Baird and Girard, Tree Lizard 

Material.— Four left maxilla (TTU-P-8334-8337); five right maxilla (TTU-P-8338-8342); 
seven left dentaries (TTU-P-8343-8349); six right dentaries (TTU-P-8350-8355). 
Discussion. — U. ornatus occurs in a wide variety of habitats in the southern Guadalupt Moun- 
tains today and is extremely common in the vicinity of Upper Sloth Cave. The wide range of 
habitats occupied by this species makes it nearly useless for climatic interpretations. 



146 



LOGAN AND BLACK 



Eumeces cf. E. multivarigatus Hallowell, Many-lined Skink 

Material. — Fragmentary left maxilla (TTU-P-8375). 

Discussion. — E. multivarigatus has previously been reported from Pratt Cave (Gehlbach and 
Holman 1974) where it represented the first record of the species from a southwestern prehis- 
toric locality. E. multivarigatus is a fairly common inhabitant in some portions of the southern 
Guadalupe Mountains and does not indicate any climatic change. 

Thamnophis sp. Fitzinger 

Material. — One precaudal vertebrae (TTU-P-8368). 

Discussion. — Thamnophis is generally considered to be primarily a mesic-adapted genus (Raun 
1965; Gehlbach and Holman 1974). The presence of this genus on the west face of the 
Guadalupe Mountains indicates either more mesic conditions or transportation of the material 
from place of origin to place of deposition. 

cf. Diadophis punctatus (Linnaeus), Northern Ringneck Snake 

Material. — One precaudal vertebrae (TTU-P-8372). 

Discussion. — D. punctatus is usually considered a woodland species (Raun 1965). This single 
specimen (Table 3) may represent transportation from the forested area approximately 450 m 
above the cave, or it may represent a depression of the woodlands to the elevation of the cave as 
suggested by Van Devender et al. (1976b) 



TABLE 3. Reptiles from Upper Sloth Cave. 








Taxa 




Depth in cm 






0-10 


10-20 20-30 


30-40 



Sceloporus poinsetti 
Sceloporus undulatus 
Sceloporus sp. 
Urosaurus ornatus 
Eumeces multivarigatus 
Eumeces sp. 
Thamnophis sp. 
Diadophis punctatus 
Coluber or Masticophis 
Opheodrys vernalis 
cf. Salvador a 
Elaphe subocularis 
Arizona elegans 
Trimorphodon biscutatus 
Crotalus sp. 
Unidentified snake 



X 


X 




X 


X 


X 


X 


X 


X 


X 


X 


X 
X 


X 

X 
X 


X 
X 


X 


X 
X 


X 


X 


X 






A 

X 


X 


X 


X 


X 


X 



Coluber Linnaeus or Masticophis Baird and Girard 

Material— Four precaudal vertebrae (TTU -P-8370); two precaudal vertebrae (TTU-P-837 1 ). 
Discussion. — We are unable to assign these vertebrae to either Coluber or Masticophis with 
certainty. M.flagellum and M. taeniatus are found in the immediate vicinity of Upper Sloth 
Cave, with the latter species being the more abundant. Brattstrom (1964) reported Coluber 
constrictor from late Wisconsin deposits of south-central New Mexico, but this species has not 



VERTEBRATE FOSSILS 147 

been reported from the southern Guadalupe Mountains. No change from present climatic con- 
ditions is indicated by this material. 

Opheodrys vernalis (Harlan), Smooth Green Snake 

Material. —Three precaudal vertebrate (TTU-P-8367). 

Discussion. — In Texas, O. vernalis presently is known only from Ellis, Bosque, Austin, and 
Matagorda counties (Raun 1965). These specimens are the first Texas records of O. vernalis 
west of the Edwards Plateau and the first confirmed record for the Guadalupe Mountains. The 
presence of O. vernalis in the southern Guadalupe Mountains has been suspected for several 
years, based on a description given to Dr. John Mecham by a rancher in the Guadalupe Moun- 
tains of southern New Mexico (John Mecham, pers. comm.) In the western United States, O. 
vernalis inhabits damp, grassy environments such as stream borders, meadows, and rocky 
habitats interspersed with grass (Stebbins 1966). The presence of this species indicates a more 
mesic environment than presently occurs in the vicinity of the cave or possibly transportation 
from place of origin. 

cf. Salvadora sp. (Baird and Girard) 

Material. — One precaudal vertebrae (TTU-P-8373). 

Discussion. — 5". grahamiae was observed within 0.5 mi. of Upper Sloth Cave during the period 

of excavation. The presence of this genus does not indicate any change in climatic conditions. 

Elaphe cf. E. subocularis (Brown), Trans-Pecos Rat Snake 

Material. — 24 precaudal vertebrae (TTU-P-8364); 10 precaudal vertebrae (TTU-P-8365); one 
precaudal vertebrae (TTU-P-8366). 

Discussion. — Raun ( 1965) listed E. subocularis from the Guadalupe Mountains and stated that 
the preferred habitat is rocky areas at higher elevations, a condition that closely matches the 
area surrounding Upper Sloth Cave. The presence of this species in the deposits does not 
indicate any change from the present climatic conditions. 

Arizona elegans Kennicott, Glossy Snake 

Material.— Two precaudal vertebrae (TTU-P-8369). 

Discussion. — A. elegans is a common inhabitant of the desert shrub community immediately to 
the southwest and approximately 500 m lower in elevation than the cave. This xeric-adapted 
species indicates no change from present climatic conditions. 

Trimorphodon biscutatus Cope, Texas Lyre Snake 

Material. — Five precaudal vertebrae (TTU-P-8374). 

Discussion. — Raun (1965) listed T. vilkinsoni{-T. biscutatus vilkinsoni) from the Trans-Pecos 
region of Texas, but excluded the Guadalupe Mountains. These specimens represent the first 
known occurrence of T. biscutatus from the Guadalupe Mountains. The preferred habitat is 
rocky, arid, or semi-arid regions where it feeds primarily on lizards (Raun 1965). No change in 
climatic conditions is indicated by this species. 

Crotalus sp. Linnaeus 

Material.— 21 precaudal vertebrae (TTU-P-8360); 13 precaudal vertebrae (TTU-P-8361); 
seven precaudal vertebrae (TTU-P-8362); one precaudal vertebrae (TTU-P-8363); two 
precaudal vertebrae (TTU-P-8377). 

Discussion. — C. atrox, C. molussus, and C. lepidus are all found in the immediate vicinity of 
Upper Sloth Cave. Based on the large size of the vertebrae, these specimens most probably 
represent C. atrox or C. molossus. Rattlesnakes frequently are found around the entrances of 
caves and their presence in the deposit was expected. Due to the wide range of habitats occupied 
by this genus, Crotalus is of little use as an ecological indicator. 



148 LOGAN AND BLACK 




Fig. 2. Recent distribution (A) and Pleistocene occurrences (x) of Sorex cinereus 
and Recent distribution (B) and Pleistocene occurrences (O) of Cryptotis parva. 



cf. Tympanuchus sp. (Linnaeus) 

Material.— One fragmentary sternum (TTU-P-8396). 

Discussion.— This fragmentary sternum closely resembles Tympanuchus sp., but is too 



VERTEBRATE FOSSILS 149 

fragmentary to permit specific identification. T. pallidicinctus presently occurs in limited num- 
bers on the plains to the east of the Guadalupe Mountains. This genus does not indicate any 
change from present climatic conditions. 

Passeriformes 

Material.— Abundant material. 

Discussion. — These bones represent at least three taxa of passeriform birds. The majority of the 
bones are of young birds and are incompletely ossified, making further identifications uncer- 
tain. 

Sorex cinereus Kerr, Masked Shrew 

Material.— Skull and Mandibles (TTI-P-8273); two, R, -M 3 (TTU-P-8274-8275); three, 
LI,-M 3 (TTU-P-8276-8278); LI'-P 4 (TTU-P-8279); LM 13 (TTU-P-8280); LM ,_ 3 
(TTU-P-8281); LP-M 2 (TTU-P-8282); RIj-M 2 (TTU-P-8283); LI,-P, (TTU-P-8284); 
LM,_ 2 (TTU-P-8285); Rl!-M, (TTU-P-8286). 

Discussion. — Sorex cinereus is differentiated from other members of the genus by a shorter and 
much shallower dentary, a shorter molar row, and alowercoronoid(Findley 1953). The closest 
occurrence of 5". cinereus today is in northern New Mexico (Fig. 2), a distance of approxi- 
mately 300 mi. Specimens from the cave deposit agree closely with a modern specimen 
( M ALB-2684) from San Miguel County, New Mexico, that is deposited in the collections of the 
Museum of Arid Lands Biology, the University of Texas at El Paso. Sorex cinereus "prefers 
mesic and hydric communities from which it rarely wanders" (Findley 1953). The presence of 
this species in the deposits is an indicator of more mesic conditions than presently occur in the 
southern Guadalupe Mountains. 

Cryptotis parva (Say), Least Shrew 

Material.— RI,-M 3 (TTU-P-8287); LI,-M 2 (TTU-P-8288); RI,-M ( (TTU-P-8289); 
LI,-P 4 (TTU-P-8290). 

Discussion. — C. parva has been reported from Dry Cave, Eddy County, New Mexico, by 
Harris et al. (1973), associated with a radiocarbon date of 10, 730±150 BP. This date compares 
favorably with the date 1 1,760±610 BP(A-1533) obtained from the 23 to 30 cm level of trench 1. 
This is the first record of Cryptotis parva from the Trans-Pecos of Texas and represents a 
former range extension of at least 200 mi. to the southwest of its present range (Fig. 2). The 
presence of C. parva in the deposits is an indicator of at least slightly more mesic conditions than 
now exist in the area. 

Notiosorex crawfordi (Coues), Desert Shrew 

Material.— Two, R^-M, (TTU-P-829 1-8292); two, LI,-M 3 (TTU-P-8293-8294); rostrum 
with RIi-M 3 , LI, -M, and M 3 (TTU-P-8295); RM,_ 2 (TTU-P-8296); LMj_ 2 
(TTU-P-8297); RM1-3 (TTU-P-8298); RP 4 -M 3 (TTU-P-8299). 

Discussion. — Specimens from Upper Sloth Cave do not differ significantly from recent speci- 
mens from Garza County, Texas. N. crawfordi possibly occurs in the vicinity of the cave today 
although no specimens are known from the Guadalupe Mountains National Park. Desert 
shrews are known from a variety of habitats and thus are relatively useless as climatic indi- 
cators. 

Myotis velifer (J. A. Allen), Cave Myotis 

Material.— LM 2 . 3 (TTU-P-8300). 

Discussion. — M. velifer is larger than other American members of the genus, with the excep- 
tion of the extinct M. magnimolaris (Choate and Hall 1967). M. velifer is differentiated from M. 
magnimolaris by a slightly less massive mandible and slightly smaller dentition; the greatest 
crown length of M. magnimolaris has a range of 1.50 to 1.65 mm and a mean of 1.57 mm 
(Choate and Hall 1967). TTU-P-8300 has a crown length of 1 .47 mm on the M 3 and is assigned 



150 LOGAN AND BLACK 

to M. velifer on this basis. M. velifer is a common inhabitant of caves in a wide variety of 
habitats and thus is not useful as an ecological indicator. 

Myotis sp. Kaup 

Material.— LP 4 -M 3 (TTU-P-8301); RP 4 -M 3 (TTU-P-8302). 

Discussion. — Several species of small Myotis are found in the immediate vicinity ol the cave 
today and are extremely difficult to differentiate on the basis of fragmentary material. The 
presence of a small Myotis in the fauna does not reflect any change in the environmental condi- 
tions. 

Lasionycteris noctivagans (Le Conte), Silver-haired Bat 

Material.— LC\-M 3 (TTU-P-8303). 

Discussion. — L. noctivagans is a migratory, tree-dwelling bat that is also known to occupy 
caves, mines, and buildings (Schwartz and Schwartz 1959). Although L. noctivagans is 
presently an uncommon bat in the Trans-Pecos, its presence in the deposits does not necessarily 
reflect any change in climatic conditions in the area. 

Eptesicus fuscus (Beauvois), Big Brown Bat 

Material.— Edentulous left mandible (TTU-P-8304); LM 2 3 (TTU-P-8305). 
Discussion. — The material from Upper Sloth Cave closely resembles a modern specimen from 
Jeff Davis County, Texas. E. fuscus commonly inhabits caves, especially in the winter, and has 
been observed hibernating in nearby caves. Its presence is not unexpected and indicates no cli- 
matic change. 

Plecotus townsendii (Cooper), Townsend's Big-eared Bat 

Material.— LM 2 3 (TTU-P-8306); RM 23 (TTU-P-8307). 

Discussion. — The above-mentioned specimens do not differ significantly from Recent speci- 
mens collected in Upper Sloth Cave. During the period of 20 July 1974 to 17 August 1974, 
Upper Sloth Cave was the site of a nursery colony of P. townsendii consisting of approximately 
50 individuals. Its presence in the deposit was not unexpected. The presence of P. townsendii 
does not indicate any change in climatic conditions. 

Antrozous pallidas (Le Conte), Pallid Bat 

Material.— LC,-M 2 (TTU-P-8308); LM 3 (TTU-P-8309); RM 2 (TTU-P-8310). 
Discussion. — A. pallidus is a rather common inhabitant of caves and mine shafts in the south- 
western United States, thus its occurrence in the deposits was expected. No change in climatic 
conditions is indicated by this species. 

Nothrotherium shastense Sinclair, Shasta Ground Sloth 

Material.— Abundant dung balls from the 25 to 45 cm level of trench 1 are referred to this 
species (TTU-P-8259). 

Discussion. — The dung of N. shastense is known from only six other North American sites — 
Rampart Cave and Mauv Caves in the Grand Canyon of Arizona; Gypsum Cave, Nevada; 
Aden Crater, New Mexico; Williams Cave and Lower Sloth Cave, Guadalupe Mountains 
National Park, Texas. Radiocarbon dates available for these locations are all 1 1,000 YBP or 
older and agree closely with the date of 1 1, 760±610 YBP (A- 1533) on artiodactyl fecal pellets 
associated with the sloth dung (Van Devender et al 1976a). No bones of N. shastense have been 
recovered from Upper Sloth Cave, but the dung balls agree closely in size and texture with the 
dung balls from the other North American sites ( Paul Martin, pers. comm.). N. shastense is the 
only extinct species represented in the deposits of Upper Sloth Cave. 



VERTEBRATE FOSSILS 151 

Sylvilagus sp. Gray 

Material.— Two, LP 3 (TTU-P-8379-8380); two, RP 3 (TTU-P-8381-8382). 
Discussion. — S. floridanus and S. auduboni both presently occur in the southern Guadalupe 
Mountains, with the latter species being more abundant. The presence of this genus gives no 
indications of climatic conditions due to the wide variety of habitats in which it is found today. 

Lepus cf. L. californicus Gray, Black-tailed Jackrabbit 

Material— LP 3 (TTU-P-8378). 

Discussion. — L. californicus is a common inhabitant of grasslands and desert areas of the 
southwestern United States (Burt and Grossenheider 1964) and is a common inhabitant of the 
shrub-desert community to the southwest of Upper Sloth Cave. This species does not indicate 
any change in climatic conditions. 

Marmota flaviventris (Audubon and Bachman), Yellow-bellied Marmot 

Material— R?t (TTU-P-8255). 

Discussion. — The present closest occurrence of M. flaviventris to Upper Sloth Cave is in the 
high mountain forests of northern New Mexico (Fig. 3). M . flaviventris has previously been 
reported from Burnet Cave (Murray 1957) and Dry Cave (Harris 1970b). Murray (1957) 
attributed the presence of this species at Burnet Cave to the movement of the forests southward 
and to a lower elevation than where they presently occur. This interpretation is supported by 
plant macrofossils and pollen samples from the southern Guadalupe Mountains (Van Devender 
et al. 1976a). In a study of a late Pleistocene fauna from north-central New Mexico, Harris and 
Findley (1964) pointed out that M. flaviventris occurs in other habitats and its presence in con- 
junction with nonforest forms, as in Dry Cave (Harris 1970b), may indicate an open habitat that 
now exists even farther to the north. Harris ( 1970a) suggested that a minimum winter rainfall of 
at least 2 in. would probably provide enough green fodder to carry M . flaviventris through the 
spring dry season. M . flaviventris is an indicator of a more mesic environment than presently 
occurs in the southern Guadalupe Mountains. 

Spermophilus variegatus (Erxleben), Rock Squirrel 

Material— Two LT^-M 3 (TTU-P-8244-8245); four, RM, or2 (TTU-P-8246-8249); two, 

LM, (TTU-P-8250, 8254); LM 1 (TTU-P-8251); RP 4 (TTU-P-8252); R and L I^-M 2 

(TTU-P-8253). 

Discussion. — The material referred to this species is indistinguishable in size and morphology 

from modern specimens from Culberson County, Texas. S. variegatus is a common inhabitant 

of rocky areas throughout the southern Gaudalupe Mountains, and has been observed in the 

immediate vicinity of Upper Sloth Cave. This species is of little value as a climatic indicator. 

Thomomys bottae (Eydoux and Gervais), Botta's Pocket Gopher 

Material— Four, LP 4 (TTU-P-8384-8386, 8390); three, RP 4 (TTU-P-8392-8393, 8387); two, 
LM l or 2 (TTU-P-8388-8389); RM lor2 (TTU-P-8391). 

Discussion. — T. bottae is the most abundant pocket gopher in the higher elevations of the 
Guadalupe Mountains today. T. bottae occupies valleys and mountain meadows of the south- 
western United States, where it prefers a loamy soil, but it also occurs in sandy or rocky soil 
(Burt and Grossenheider 1964). This species is of little value as a climatic indicator. 

Pappogeomyscastanops(Baird), Yellow-faced Pocket Gopher 

Material— LP 4 (TTU-P-8383). 

Discussion.— P. castanops is found primarily on the high plains and in mountain basins (Blair 
et al. 1957) and is found in the Guadalupe Mountains today. This species is of little use as a cli- 
matic indicator. 



152 LOGAN AND BLACK 




Fig. 3. Recent distribution (A) and Pleistocene occurrences (x) of Marmota 
flaviventris. 



VERTEBRATE FOSSILS 153 




Fig. 4. Recent distribution (A) of Neotoma cinerea and present area of sympatry (B) 
for Sorex cinereus, Cryptotis parva, Marmota flaviventris, and Neotoma cinerea. 



154 LOGAN AND BLACK 

Peromyscus spp. Gloger 

Material.— Abundant fragmentary material. 

Discussion. — P. eremicus, P. leucopus, P. maniculatus, P. boylei, P. truei, P. difficilis, and P. 
pectoralis all presently occur in the southern Guadalupe Mountains (Genoways et al. 1977). Of 
these seven species, only P. eremicus can be identified to species with any degree of certainty on 
the basis of fragmentary material. P. eremicus is differentiated from the other six species by the 
lack of, or, at the most, rudimentary accessory cusps in the two principal outer angles of the M 1 
and M 2 (Hall and Kelson 1959). Due to the extreme range in habitats occupied by members of 
this genera and the difficulty of distinguishing the various species, Peromyscus are nearly use- 
less as climatic indicators. 

Peromyscus eremicus (Baird), Cactus Mouse 

Material— RM 1 2 (TTU-P-8394); RM 1 (TTU-P-8395). 

Discussion. — Characters to differentiate P. eremicus have already been discussed in the pre- 
vious account of Peromyscus spp. P. eremicus is an inhabitant of deserts from central Mexico 
through the southwestern United States (Blair et al. 1957). The presence of this species in the 
fauna does not indicate any change from present climatic conditions. 

Neotoma cinerea (Ord), Bushy-tailed Woodrat 

Material.— RM 1 (TTU-P-8264); RM 2 (TTU-P-8265); LM, (TTU-P-8266); three, LM 2 
(TTU-P-8267-8270); RM 2 (TTU-P-8271); LMi_ 2 (TTU-P-8272). 

Discussion. — N. cinerea is differentiated from N. mexicana on the basis of accessory cusps 
developed in the re-entrant angles of some of the teeth (Lundelius 1976), a condition found in 
50% of the Recent specimens of N. cinerea examined and lacking in all Recent specimens of N. 
mexicana examined. 

N. cinerea is essentially a boreal animal and is found at higher latitude or higher elevations 
today (Hall and Kelson 1959). The present closest population of N. cinerea to Upper Sloth Cave 
is in the mountains of north-central New Mexico (Fig. 4). The presence of N. cinerea in the 
fauna is an indicator of cooler and /or more mesic conditions than now exist in the southern 
Guadalupe Mountains. 

Neotoma mexicana Baird, Mexican Woodrat 

Material. — Abundant isolated teeth from all levels of trench 1 and trench 2. 
Discussion.— N. mexicana is identified on the basis of dentine tracts on the anteroexternal sides 
of the M i that extend from one-fourth to one-third the distance from the root to the crown of an 
unworn tooth. The dentine tracts on the M 2 are shorter (Lundelius 1977). 

This species is very common among the limestone ledges and cliff faces that dominate the west 
face of the Guadalupe Mountains. N. mexicana is not useful as a climatic indicator. 

Neotoma micropus (Hartly), Southern Plains Woodrat 

Material,— LM\ (TTU-P-8400). 

Discussion. — Dalquest et al. ( 1 969) separated N. micropus from N. albigula on the width of the 
second lophid of the M i . They found that this measurement in TV. albigula was always less than 
1.94 mm whereas in N. micropus this measurement was always greater than 1.94 mm. This 
criterion was followed in the identification of the Neotoma from Upper Sloth Cave. When these 
two species are sympatric, N. albigula is restricted to rocky areas and N. micropus is restricted to 
more open areas (Finley 1958). N. micropus presently occurs in the flats in the western portion 
of the park. 

Neotoma albigula (Baird), White-throated Woodrat 

Material.— Two, LM, (TTU-P-840 1-8402); RMi (TTU-P-8403). 

Discussion.— Criteria for identification and habitat preferences were discussed in the preced- 



VERTEBRATE FOSSILS 155 

ing account of N. micropus. N. albigula occurs at middle to lower elevations in the park at 
present. 

Microtus mexicanus (Saussure), Mexican Vole 

Material. — Abundant isolated teeth. 

Discussion. — M. mexicanus presently occurs in grassy meadows in the higher elevations of the 
southern Guadalupe Mountains and is common in local areas. The presence of this species in 
the fauna does not indicate any significant change in climatic conditions. 

Bassariscus astutus (Lichtenstein), Ringtail 

Material.— TP 3 (TTU-P-8256); LM, (TTU-P-8257). 

Discussion.— B. astutus is an inhabitant of the more rocky areas of the southern Guadalupe 
Mountains where it feeds on a wide variety of small mammals, birds, insects, and plants. This 
species does not indicate any change in climatic conditions. 

Mustela frenata (Lichtenstein), Long-tailed Weasel 

Material.— RP3-M3 (TTU-P-8258). 

Discussion. — M. frenata has not been taken from the Guadalupe Mountains National Park in 
Recent times (Genoways et al. 1 977) although it occurs widely in the United States and Mexico. 
It has been recorded from Culberson County (Davis 1966:87). M. frenata is found in a variety of 
habitats and is therefore not useful as an ecological indicator. 

Felis concolor Linnaeus, Mountain Lion 

Material.— RdP 3 (TTU-P-831 1). 

Discussion. — F. concolor presently occurs in limited numbers in the southern Guadalupe 
Mountains. The presence of this species in the deposit was not unexpected and gives no infor- 
mation concerning climatic conditions. 



CLIMATIC INTERPRETATIONS 

Upper Sloth Cave provides a good record of faunal transition from a 
more mesically adapted vertebrate community to the present more xerical- 
ly adapted vertebrate community. Five taxa of xerically adapted 
vertebrates are found in the to 10 cm level of trench 2, whereas only two 
taxa of mesically adapted vertebrates occur in this level. The 10 to 20 cm 
level of trench 2 presents the opposite picture, with five taxa of mesically 
adapted vertebrates and only two taxa of xerically adapted vertebrates still 
present in the deposit. In the 20 to 30 cm level, one additional taxon of 
mesically adapted vertebrate is gained, whereas all xerically adapted verte- 
brates are absent. This drop in the number of mesically adapted taxa can 
probably be attributed to the relatively small amount of matrix removed 
from the 20 to 30 level of trench 2 because of the configuration of the cave 
walls at this depth. The actual numbers of specimens obtained also follow 
the trends discussed above. 

If we assume that Sorex cinereus, Cryptotis parva, Marmotaflaviventris, 
and Neotoma cinerea existed contemporaneously in the vicinity of Upper 
Sloth Cave and look for a modern area of sympatry, we find that the 
eastern portion of the Black Hills of South Dakota is the only area where 
these four taxa presently occur together. The elevation of the Black Hills is 



156 LOGAN AND BLACK 

similar to that of Upper Sloth Cave and the vegetation of the Black Hills is 
similar to that proposed for the southern Guadalupe Mountains by Van 
Devender et al. (1976b). On the basis of the previously discussed floral and 
faunal similarities, we are therefore postulating that approximately 1 1,000 
years ago the climatic conditions in the southern Guadalupe Mountains 
may have been similar to the climatic conditions existing in the Black Hills 
of South Dakota today. 

This more xeric trend in the southern Guadalupe Mountains may have 
begun approximately 11,500 years ago and been a contributing factor in 
the extinction of the Shasta Ground Sloth, Nothrotherium shastense, or it 
may not have started until as late as 5000 years ago as suggested by Van 
Devender and Worthington (1977). Lack of a radiocarbon date directly 
associated with the deposits in trench 2 precludes a more definite timetable. 



LITERATURE CITED 

Ayer, M. W. 1936. The archaeological and faunal material from Williams Cave, 

Guadalupe Mountains, Texas. Proc. Acad. Nat. Sci. Phila. 78:599-618. 
Blair, W. F., A. P. Blair, P. Brodkorb, F. R. Cagel and G. A. Moore. 1957. 

Vertebrates of the United States. McGraw-Hill Book Co., Inc., New York, 8 19 pp. 
Brattstrom, B. H. 1964. Amphibians and reptiles from cave deposits in south- 
central New Mexico. Bull. South. Calif. Acad. Sci. 63:93-103. 
Burt, W. H., and R. P. Grossenheider. 1964. A Field Guide to the Mammals. 

Houghton Mifflin Co., Boston, 2nd ed., 284 pp. 
Choate, J. R., and E. R. Hall. 1967. Two new species of bats, genus Myotis, 

from a Pleistocene deposit in Texas. Amer. Midi. Nat. 78:531-534. 
Dalquest, W. W., E. Roth, and F. Judd. 1969. The mammal fauna of Schultze 

Cave, Edwards County, Texas. Bull. Fla. State Mus. Biol. Ser. 13:205-276. 
Davis, W. B. 1966. The mammals of Texas. Bull. Tex. Parks Wildl. Dep. 41:1-267. 
Findley, J. S. 1953. Pleistocene Soricidae from San Josecito Cave, Nuevo Leon, 

Mexico. Univ. Kans. Publ. Mus. Nat. Hist. 5:635-639. 
Findley, R. B., Jr. 1958. The wood rats of Colorado: distribution and ecology. 

Univ. Kans. Publ. Mus. Nat. Hist. 10:213-552. 
Gehlbach, F. R., and J. A. Holman. 1974. Paleoecology of amphibians and 

reptiles from Pratt Cave, Guadalupe Mountains National Park, Texas. South- 
west. Nat. 19:191-198. 
Genoways, H. H., R. J. Baker, and J. E. Cornely. 1977. Mammals of the 

Guadalupe Mountains National Park, Texas. This volume. 
Hall, E. R., and K. R. Kelson. 1959. The Mammals of North America. Ronald 

Press, New York. 2 vols. 
Harris, A. H. 1970a. Past climate of the Navajo Reservoir District. Am. 

Antiquity 35:374-377. 
1970b. The Dry Cave mammalian fauna and late pluvial conditions in 

southeastern New Mexico. Tex. J. Sci. 12:3-27. 
Harris. A. H., and J. S. Findley. 1964. Pleistocene-Recent fauna of the Isleta 

Caves, Bernalillo County, New Mexico. Am. J. Sci. 262:114-120. 
Harris, A. H., R. A. Smartt, and W. R. Smartt. 1973. Cryptotis parva from 

the Pleistocene of New Mexico. J. Mammal. 54:512-513. 



VERTEBRATE FOSSILS 1 57 

Holm an, J. A. 1970. A Pleistocene herptofauna from Eddy County, New Mexico. 
Tex. J. Sci. 22:29-39. 

Logan, L. E. 1975. The Quaternary vertebrate fauna of Upper Sloth Cave, Guada- 
lupe Mountains National Park, Texas. Geol. Soc. Am. Southcentral Sec. p. 210 
(abstract). 

Lundelius, E. L., Jr. 1977. Post-Pleistocene mammals from Pratt Cave and their 
environmental significance. This volume. 

Mera, H. P. 1938. Reconnaissance and excavation in southeastern New Mexico. 
Mem. Am. Anthropol. Assn. 51:34-36. 

MURRAY, K. F. 1957. Pleistocene climate and the fauna of Burnet Cave, New Mex- 
ico. Ecology 38:129-132. 

Raun, G. G. 1965. A guide to Texas snakes. Mus. Notes Tex. Mem. Mus. 9:1-85. 

Schultz, C. B., and E. B. Howard. 1935. The fauna of Burnet Cave, Guadalupe 
Mountains, New Mexico. Proc. Acad. Nat. Sci. Phila. 87:273-298. 

Schwartz, C. W., and E. R. Schwartz. 1959. The Wild Mammals of Missouri. 
Univ. Missouri Press and Missouri Conserv. Comm., 341 pp. 

Stearns, C. E. 1942. A fossil marmot from New Mexico and its climatic sig- 
nificance. Am. J. Sci. 240:867-878. 

Stebbins, R. C. 1966. A Field Guide to Western Reptiles and Amphibians: Field 
Marks of All Species in Western North America. Houghton Mifflin Co., Boston, 
279 pp. 

Van Devender, T. R., and R. D. Worthington. 1977. The herptofauna of 
Howell's Ridge Cave and the paleoecology to the northwestern Chihuahuan 
Desert. In R. H. Wauer and D. H. Riskind, eds. Transactions Symposium on the 
Biological Resources of the Chihuahuan Desert Region, U.S. and Mexico, Na- 
tional Park Service, Washington, D.C. 

Van Devender, T. R., P. S. Martin, A. M. Phillips III, and W. G. 
Spaulding. 1977a. Late Pleistocene biotic communities from the Guadalupe 
Mountains, Culberson County, Texas. In R. H. Wauer and D. H. Riskind, 
eds.Transctions Symposium on the Biological Resouces of the Chihuahuan 
Desert Region, U.S. and Mexico, National Park Service, Washington, D.C. 

Van Devender, T. R., W. G. Spaulding, and A. M. Phillips III. 1977b. Late 
Pleistocene plant communities in the Guadalupe Mountains, Culberson County, 
Texas. This volume. 



ACKNOWLEDGMENTS 

We would like to thank Mr. Donald Dayton, Superintendent of Guada- 
lupe Mountains National Park and the National Park Service for permis- 
sion to excavate Upper Sloth Cave. Dr. Gary Ahlstrand, John Chapman, 
Phil Van Cleave, and Charlie Peterson of the National Park Service assisted 
in many ways. We would also like to extend a special thanks to Roger Reisch 
for his invaluable friendship and assistance in all phases of this project. 
Richard Fullington, Dallas Museum of Natural History, and Hal Pierce, 
Department of Geosciences, Texas Tech University, identified the 



158 LOGAN AND BLACK 

molluscan fauna. Dr. Thomas Van Devender, University of Arizona, 
assisted in some of the reptile identifications as well as identifying all the 
plant macrofossils. Dr. Paul Martin, University of Arizona, provided all 
radiocarbon dates as well as the pollen analysis and many valuable hours of 
discussion. Drs. Ernest Lundelius, University of Texas at Austin, Arthur 
Harris, University of Texas at El Paso, and Hugh Genoways, Texas Tech 
University, provided access to collections in their care as well as partici- 
pating in many hours of discussion. Albert Hernandez sorted the bulk of the 
screened concentrate. Dave Bohaska provided invaluable assistance during 
the excavations. We would also like to thank the many others, too numerous 
to mention by name, who have contributed to the completion of this project. 



Environmental Implications of 
Herpetofaunal Remains from 
Archeological Sites West of 
Carlsbad, New Mexico 



JOHN S. APPLEGARTH, University of New Mexico, 
Albuquerque 

This report is to demonstrate the usefulness of herpetofaunal remains in 
reconstructing archeological environments. In and near caves and rock- 
shelters predators leave a concentrated sample of whatever prey species are 
abundant locally. Three such sites in the hills west of Carlsbad, Eddy 
County, New Mexico, have yielded a wealth of faunal remains, and among 
the smallest bones are those of amphibians and reptiles. 

In the arid Southwest the archeological remains of salamanders, turtles, 
and snakes are relatively uninformative. The tiger salamander, Ambystoma 
tigrinum, is widespread and only suggests general proximity to a pond or 
streamcourse. Finding any other species of salamander is unlikely. Turtle 
remains in archeological sites are environmentally unreliable because the 
human residents may have carried them great distances for uses other than 
food. Snakes often make up a major part of archeological herpetofauna; 
unfortunately most species are relatively wide-ranging, so their remains say 
little about local conditions. Accordingly, this report is limited to the two 
groups that seem most useful as indicators of past conditions, the anurans 
(frogs and toads) and the lizards. 

The oldest of the three sites is Honest Injun Cave, which is located on the 
West Fork of the Little McKittrick Draw at 1 1 15 m (3660 ft) elevation (Fig. 
1). It is situated about 12 m up a terraced, west-facing, limestone cliff, 
extends in horizontally for about 12 m, and has about 60 cm of loose dust 
and rock debris on the floor. A radiocarbon date of 2930 ± 60 years BP (980 
B.C.; WIS-598) was obtained from woven artifacts found near the middle of 
the cave and about 25 cm below the surface. Depthwise, the artifacts tended 
to be concentrated in the middle third, but occupants had churned the fill to 
some extent. Three carapace fragments of a western box turtle, Terrapene 
ornata, all from a single individual judging by color and size, had been 
scattered at least 15 cm vertically and 3 m horizontally. Roughly 100 m 

159 



160 



APPLEGARTH 




FIG. 1. Map of the area west of Carlsbad, New Mexico, showing the distributions of 
the following archeological sites (shaded dots) and paleontological sites (crosses): 
Ellis, Roberts Rockshelter, Dry Cave, Honest Injun Cave, and Dark Canyon Cave 
(reading from upper left to lower right). The elevational contours, abstracted from 
USGS maps, are in feet above mean sea level. 



downcanyon is an arroyo-bottom pool; although never seen completely dry, 
its water is foul when low. Except for the Little McKittrick floodplain, the 
surrounding area within 5 km is almost entirely rolling, rocky hills. 

The second site is Roberts Rockshelter located on Dunnaway Draw at 
1183 m (3880 ft) elevation. The shelter is situated in the lower part of a west- 
facing, limestone cliff which is about 8 m high. It is adjacent to an arroyo- 
bottom pool, which usually has some water but has been seen completely 
dry. A radiocarbon date of 1075 ± 50 years BP (A.D. 875; WIS-579) was 
based on charcoal found 90 cm down in the midden adjacent to the shelter. 
The deposit was entirely outside the shelter, had a maximum depth of about 
1 10 cm next to the shelter, and became thinner as one moved away from the 
cliff, with bare rock benches rimming the pool. Raptors have used the shelter 
and nearby ledges as evidenced by the thousands of small bones recovered 
from roughly a cubic meter of deposit, and the deposit being richest within 
one meter of the cliff. A similar number of bones was recovered from deposit 
disturbed by illegal artifact hunters (the "back dirt" of "pot hunters"). 

Third, the Ellis site consists of two small shelters in a low, west-facing bluff 
on the south side of Rocky Arroyo at 1 1 73 m (3850 ft) elevation. The shelters 
are bo*h about 1 m high and extend horizontally 4 m and 2 m, respectively, 
into the loosely cemented conglomerate just under a relatively hard top 



HERPETOFAUNAL REMAINS 161 

stratum. Radiocarbon dates of 8 15 ±50 years BP(A.D. 1135; WIS-577) and 
8 10 ± 55 years BP ( A.D. 1 140; WIS-578) were obtained from charcoal found 
23 cm and 45 cm below the surface of the midden deposit just outside each 
shelter. The fill in the shelters was 50 cm and 30 cm deep. Test trenches were 
dug in the fill and extended outward through the midden of each shelter. As 
the shelters were only 51 m apart and so similar in age and design, their 
herpetofaunal data are pooled in this report. Occupants had churned the fill 
as evidenced by two bones of the barking frog, Hylactophryne augusti, evi- 
dently from one individual, one bone being on the surface and the other at 
least 28 cm down in the 50-cm fill. Both the Roberts and Ellis sites are in a 
shallow valley, Indian Basin, the topography of which is a mixture of rela- 
tively level areas and low, rolling hills. 

Because the deposits of all the sites were more or less disturbed, strati- 
graphic data are not presented. In treating data for whole sites, it is assumed 
that the proportion of bones from historic times is relatively insignificant. 
Taking relative recoverability into account, the archeological samples are 
viewed as a fair measure of the prehistoric relative abundance of the various 
species, before modern environmental alteration. 

For comparison with archeological abundance, qualitative estimates of 
present relative abundance (PR A) within 5 km of each site are given with the 
archeological data in Tables 1 and 2. These estimates are based on four 
summers of field work (1970-73) and the museum records of Carlsbad 
Caverns National Park, Eastern New Mexico University, the University of 
Kansas, and the University of New Mexico. All archeological herpetofauna 
is being held for further study and eventually will be deposited in the Labora- 
tory for Environmental Biology of the University of Texas at El Paso. 

Of the anuran and lizard species now in the general area but not found in 
the archeological sites, four are so small that recovery of their delicate bones 
was unlikely with the single, dry, flyscreen sieves (1 to 1.5-mm mesh) that 
were used (layered wet screens, down to 0.5-mm mesh, would have been 
better). The four species are: the cricket frog, Acris crepitans (which, because 
of its limitation to the immediate vicinity of permanent water, was not seen 
or heard near any of the sites); the green toad, Bufo debilis; the tree lizard (a 
misnomer in New Mexico as most are found on large rocks), Urosaurus 
ornatus; and the many-lined skink, Eumeces multivirgatus . The last two are 
scarce and most often found near springs and permanent streams. 

The only relatively large species ' not found in the sites was the bullfrog, 
Rana catesbeiana, which is suspected to have invaded the Pecos River in 
historic times (Raun and Gehlbach 1972:10). Although this species was 
reported by Wiley (1972) from a Pleistocene deposit in Dark Canyon Cave 
(Fig. 1), personal reexamination of the material negates that record. 



'The plains leopard frog, Rana blairi, may inhabit the Pecos Valley as far south as Artesia, 
roughly 43 km NNE of the Ellis site; it seems to be separable osteologically from the Rio Grande 
leopard frog, R. berlandieri. 



162 APPLEGARTH 

University of Texas 41228-509 consists of two third vertebrae, one from a 
large (roughly 1 1 cm) Woodhouse's toad, Bufo woodhousei, and the second 
from a smaller (roughly 7 cm) toad, Bufo sp. (possibly B. woodhousei, but 
the bone is in poor condition). 

ANURANS 

The red-spotted toad, Bufo punctatus (Table 1), locally prefers rocky 
slopes. The rocky terrain around Honest Injun explains both its archeo- 
logical and modern abundance. Although now common in the Indian 
Basin, it seems to have been relatively scarce there 800 years ago, a change 
presumably due to overgrazing and consequent soil erosion. 

The Texas toad, Bufo speciosus, is fairly restricted to valleys with some 
grass and soil. In the Indian Basin its archeological abundance contrasted to 
its present-day rarity suggests considerable erosion and vegetational altera- 
tion. The relative abundance of B. speciosus compared to B. punctatus, 
wherever sympatric, seems a sensitive indicator of the relative abundance of 
grass and soil. (Many bones could only be assigned to species groups, so 
when some were determined to a particular species and none to other closely 
related species, all were pooled for estimating the minimum number of 
individuals represented.) 

Woodhouse's toad, Bufo woodhousei, is now absent from Eddy County, 
with the exception of the Pecos floodplain from Lake McMillan northward 
(25 airline km northeast of Ellis). It favors sandy soil, relatively mesic condi- 
tions, and is now generally confined to river valleys in the Southwest. It has 
been reported from Pleistocene deposits (Fig. 1) in Dry Cave(Holman 1970) 
and Dark Canyon Cave (Wiley 1972). The remains in Honest Injun Cave 



TABLE 1. Archeological and present anuran faunas of three sites west of Carlsbad, New 
Mexico. Archeological items include fragments, whole bones, and groups of co-adherent 
bones. MNI are the minimum number of individuals represented by the recovered items. 
Estimates of present relative abundance (PR A) within 5 km of each site are very common 
(vc), common (c), scarce (s), rare (r), and absent (a). 



Anuran species 


Honest Injun 




Roberts 






Ellis 




Items 


MNI 


PRA 


Items 


MNI 


PRA 


Items 


MNI 


PRA 


Bufo debilis 


\ 




s 






c 


\ 




c 


Bufo punctatus 


51 


10 


vc 


9 


3 


c 


4 } 


2 


c 


Bufo small species 


4J 






\ 






2J 






Bufo speciosus 






a 


9 


10 


r 






s 


Bufo large species 


,7 1 






67 j 






7 


3 




Bufo woodhousei 


6 ) 


6 


a 






a 






a 


Hylactophryne augusti 


r 


1 


a 


\ 




a 


2 


1 


a 


Rana berlandieri 






r 


l) 


2 


r 






r 


Rana small species 








1J 












Scaphiopus couchi 


2 


1 


a 


161 


17 


r 


46 


8 


s 


Spea ham.nondi 






c 


30} 


23 


c 


'} 


2 


c 


Spea species 


4 


1 




196 J 






»j 







HERPETOFAUNAL REMAINS 163 

suggest a slightly wetter climate and locally marshy conditions along the 
Little McKittrick Draw, probably in post- Wisconsin times and possibly as 
recent as 3000 years ago. The absence of this species from the Indian Basin 
deposits indicates attainment of the present, relatively dry climate prior to 
1000 years ago. 

The barking frog, Hylactophryne augusti, is rare in New Mexico. Of the 
five modern specimens (all in the University of New Mexico herpetology 
museum), four are from east of Roswell and one is from 1 9 airline km north- 
west of Carlsbad. All were collected on grassy benches adjacent to cliffs 
along the Pecos River. Two archeological bones from Ellis and one from 
Honest Injun suggest some grass on the benchland above each bluff. More 
important to their distribution seems to be their ability to find moisture in 
limestone cracks and caves, and their apparent near extinction in the 
Carlsbad area suggests a lowering of the ground moisture. 

The Rio Grande leopard frog, Rana berlandieri, is a member of the 
recently redefined Rana pipiens complex (Pace 1974). All species of leopard 
frogs indicate permanent water somewhere in the general vicinity; this spe- 
cies may indicate a relatively warm climate. In 1973 four live frogs at the 
Honest Injun pool and two at the Roberts pool were a surprise because none 
had been seen there in three previous summers. These vagrants probably 
moved in during a rainy period; their presence seemed ill-fated, as most bio- 
logical range-expanding efforts must be. The source of the Honest Injun 
frogs is unclear, possibly a spring-refugium somewhere upcanyon. The frogs 
at the Roberts pool most likely came from spring-fed pools about 10 km 
downcanyon. In general, ranid frogs breed only in permanent water, where- 
as spadefoot toads breed only in temporary water. Honest Injun yielded no 
ranid bones and six spadefoot bones to suggest ephemerality of the local 
pool. The ephemerality of the Roberts pool is certified by a scarcity of ranid 
remains, two bones, and an abundance of spadefoot remains, 560 bones, 
which were found at all depths in the deposit. Arroyo-bottom and rain pools 
around Ellis are highly ephemeral, and that seems to have been the case for at 
least the last 800 years. 

Couch's spadefoot toad, Scaphiopus couchi, prefers short-grass and 
desert grasslands with deep, sandy alluvium. It is now common in the rela- 
tively level parts of the Pecos Valley but seems to be absent from the Honest 
Injun area and scarce in the Indian Basin. Its archeological presence in 
Honest Injun suggests the occurrence of some grassland in the vicinity, most 
likely on the Little McKittrick floodplain. In the Indian Basin the archeo- 
logical abundance of this species is evidence for a relatively extensive grass- 
land and deep soil prior to modern grazing. 

The western spadefoot toad, Sped 1 hammondi, locally appears to prefer 



2 Most authors have treated Spea as a subgenus or synonym of Scaphiopus; Spea is used here 
as a valid genus because most skeletal elements can be determined to either Spea or Scaphiopus 
(a practical matter in working with herpetofaunal remains), and because the magnitude of the 
osteological differences is viewed as justification for full generic rank. 



164 APPLEGARTH 

hilly terrain with sandy or gravelly substrate. However this may be a margi- 
nal habitat, unsuitable for the larger S. couchi which may displace S. 
hammondi from the Pecos Valley. The greater abundance of S. hammondi 
bones in Roberts as compared to Ellis probably reflects the terrain around 
each site, Ellis having relatively more flat land. The abundance of both S. 
couchi and S. hammondi in the Indian Basin sites indicates, in prehistoric 
times, both grassland in the relatively level areas and open desert vegetation 
on the hillsides. 

LIZARDS 

Whiptail lizards, Cnemidophorus sp. (Table 2), of this region include five 
species. Arranged by body size from smallest to largest, they are C. 
inornatus, C. gularis, C. exsanguis, C. tigris, and C. tesselatus. Lack of ade- 
quate comparative material has prevented specific identification of remains. 
However, with considerable overlap among the ranges of these species, use 
can be made of the observation by Asplund (1974:695) that "the larger 
species tend to be restricted to habitats that are shaded, relative to the open 
habitats of smaller species." Of the minimum of four individuals repre- 
sented by remains from Honest Injun, two were small and two were medium 
to large; of the eight from Roberts, four were small; and of the four from 
Ellis, three were small. Presence of medium to large species in all three sites 
indicates the occurrence of some brush, sotol, or other shade-providing 
vegetation in the general vicinity of all three sites. 

The Texas banded gecko, Coleonyx brevis, is found in arid regions where 
there are fractured rock outcrops. The apparent low frequency of remains is 
probably due largely to the difficulty of recovering their delicate bones. 



TABLE 2. Archeological and present lizard fauna of three sites west of Carlsbad, New Mexico. 
Abbreviations as in Table 1. 



Lizard species 


Honest Injun 




Roberts 






Ellis 




Items 


MNI 


PRA 


Items 


MNI 


PRA 


Items 


MNI 


PRA 


Cnemidophorus species 


9 


4 


vc 


31 


8 


vc 


7 


4 


vc 


Coleonyx brevis 






s 






s 


1 


1 


s 


Crotaphytus collaris 


28 


5 


c 


77 


13 


c 


8 


2 


c 


Eumeces obsoletus 


6 


2 


r 


15 


4 


r 






r 


Holbrookia texana 


5 


3 


c 


29 


11 


c 


1 


1 


c 


Phrynosoma cornutum 


6 


2 


a 


39 


6 


r 


7 


2 


r 


Phrynosoma douglassi 


14 


4 


a 






a 






a 


Phrynosoma modestum 


4, 


1 


c 


278 


47 


c 


6 


3 


c 


Sceloporus poinsetti 


58} 


16 


c 


44 \ 


8 


s 






r 


Sceloporus large species 


7 J 






5 J 












Sceloporus undulatus 


11 


5 


vc 


26 


7 


c 


16 


6 


vc 


Urosaurus ornatus 






r 






r 






a 


Uta stansburiana 






a 


1 


1 


r 






r 



HERPETOFAUNAL REMAINS 165 

The collared lizard, Crotaphytus collaris, ranges from hot desert to 
pinyon-juniper woodland, wherever it can find rocks for lookouts and open 
areas for running. As noted by Gehlbach and Holman (1974:196), this 
species may have been in the aboriginal diet. Consequently their remains are 
relatively uninformative. 

The Great Plains skink, Eumeces obsoletus, is relatively riparian, and 
some riparian vegetation is suggested by their remains in Honest Injun and 
Roberts. Their absence from Ellis could be due to sampling error, as the 
small deposits yielded a minimum of only 19 lizards. 

The greater earless lizard, Holbrookia texana, prefers open stands of 
desert vegetation with plenty of bare ground for running. On study plots in 
Big Bend National Park, Degenhardt (1977) found this species was elimi- 
nated by dense growth of grass where the ground formerly had been rela- 
tively bare due to grazing. The abundance of their remains in Roberts and 
presence in the other sites indicate an open nature of some portion of the 
local vegetation, especially near Roberts. The open desert vegetation of the 
hill slopes probably has changed little in historic times. 

The Texas horned lizard, Phrynosoma cornutum, favors grasslands with a 
relatively warm climate and soft, sandy soil. It is also found in disturbed 
areas that used to be grassland. It is fairly common in the Pecos Vailey, but a 
specimen from 1 1 km east of Ellis is the closest modern record to the Indian 
Basin 3 . Its archeological presence in all three sites indicates some grassland 
with soft soil to have been near all three sites. 

The short-horned lizard, Phrynosoma douglassi, favors grasslands with 
loose soil but with cooler and wetter climatic conditions. Its range extends 
north into Canada where it inhabits the valleys between mountain ranges, 
and south into Mexico where it is confined to the higher elevations of the 
Sierra Madre Occidental. In the Guadalupe Mountains of Texas and New 
Mexico this species is scarce and only known from above 1800 m elevation. 
It is also known from Pleistocene deposits in Dry and Dark Canyon caves at 
1280 m and 1067 m elevation, respectively (Holman 1970; Wiley 1972). 
Therefore the remains in Honest Injun Cave indicate a slightly cooler and 
wetter climate, probably post-Wisconsin and possibly as recent as 3000 years 
ago. The absence of this species from the Indian Basin deposits indicates 
warmer, drier times by 1000 years ago. 

The round-tailed horned lizard, Phrynosoma modes turn, prefers a warm 
climate and a substrate of loose sand or gravel, with little or no grass. It is 
now common on the hills west of Carlsbad but seems generally absent from 
the Pecos floodplain, which is in direct contrast to the distribution of P. 
cornutum. The few bones in Honest Injun and Ellis suggest some loose 



3 In June 1976, a specimen (UNM 31935) was found in the Indian Basin, dead on state road 
137 about 1.3 km southeast of the Roberts site. 



166 APPLEGARTH 

gravel and open desert vegetation locally, but these bones could be of rela- 
tively recent origin. The wealth of remains in Roberts probably represents 
concentration via raptors and their pellets. A ratio of 6 P. cornutum to 47 P. 
modestum indicates strongly that the hills around Roberts were much as 
they are today, with loose gravel and open desert vegetation, and that the 
prehistoric grassland was relatively limited to the floodplains. 

Not only are the massive skull bones of the horned lizards relatively 
durable and recoverable, but they are also highly identifiable to species. 
Therefore the genus Phrynosoma is particularly useful as an indicator of 
past conditions. The three species of this region form a gradient of environ- 
mental preferences with P. douglassi favoring moderately cool, mesic grass- 
land, P. cornutum preferring warmer, drier grassland, and P. modestum 
favoring desert with little or no grass. 

The crevice spiny lizard, Sceloporus poinsetti, is generally restricted to 
outcrops of large, fractured rock, individuals usually remaining close to a 
crevice retreat. The relative number of their bones in the three sites 
corresponds well with the relative abundance of nearby suitable habitat. 

The eastern fence lizard, Sceloporus undulatus, locally favors dense plants 
such as yucca, agave and sotol, especially in association with loose rocks and 
crevices. The ratio between S. poinsetti and S. undulatus in Honest Injun is 
biased by the greater recoverability of the bones of the larger species, S. 
poinsetti. Both species now live in the vicinity of Honest Injun, but S. 
undulatus is far more abundant. The quantity of S. undulatus remains from 
the Ellis shelters reflects its modern abundance along the bluff and suggests 
the presence of bushes and other dense plants on the bluff face in prehistoric 
times. 

The side-blotched lizard, Uta stansburiana, locally seems to prefer loose 
sand and open vegetation. It is common in the sand hills east of the Pecos, 
less common on old sandbars of the Pecos riverbed, and a single specimen 
from 3 km east of Ellis seems to approximate the present limit of its distribu- 
tion into the limestone hills west of the Pecos. The single bone from Roberts 
could be of relatively recent origin, and its singularity indicates the local area 
to have been marginal or unsuitable for this species in prehistoric times. 

SUMMARY 

Species-by-species discussion of the anurans and lizards represented in 
three archeological sites in southeastern New Mexico, paints a detailed 
picture of environmental conditions as they were in prehistoric times, prior 
to overgrazing of delicate desert grassland by domestic livestock. 

LITERATURE CITED 

Asplund, K. K. 1974. Body size and habitat utilization in whiptail lizards 

{Cnemidophorus). Copeia 1974:695-703. 
Degenhardt, W. G. 1977. A changing environment — documentation of lizards 

and plants over a decade. In R. H. Wauer and D. H. Riskind, eds. Symposium 



HERPETOFAUNAL REMAINS 167 

on the Biological Resources of the Chihuahuan Desert Region, United States and 
Mexico. National Park Service, Transaction-Symposium Series No. 3, Govern- 
ment Printing Office, Washington, D.C. 

Gehlbach, F. R., and J. A. Holman. 1974. Paleoecology of amphibians and rep- 
tiles from Pratt Cave, Guadalupe Mountains National Park, Texas. Southwest. 
Nat. 19:191-197. 

Holman, J. A. 1970. A Pleistocene herpetofauna from Eddy County, New Mex- 
ico. Tex. J. Set 22:29-39. 

Pace, A. E. 1974. Systematic and biological studies of the leopard frogs (Rana 
pipiens complex) of the United States. University of Michigan Museum of Zool- 
ogy Miscellaneous Publications 148:1-40. 

Raun, G. G., and F. R. Gehlbach. 1972. Amphibians and reptiles in Texas. 
Dallas Mus. Nat. Hist. Bull. 2:1-132. 

Wiley, E. O. 1972. The Pleistocene herpetofauna of Dark Canyon Cave, New 
Mexico. Herpetol. Rev. 4:128. 



ACKNOWLEDGMENTS 

For the loan of comparative skeletons of several scarce species I am grate- 
ful to J. Alan Holman of Michigan State University. I would like to thank 
Barry Hinderstein of Sam Houston State University and Ernest L. 
Lundelius of the University of Texas at Austin for the loan of herpeto- 
faunal remains from Dark Canyon Cave. W. H. "Bill" Balgemann, Sr., 
William G. Degenhardt, Arthur H. Harris, J. Alan Holman, and Dean Ricer 
have generously provided facilities, assistance, and/ or advice. Vital to this 
work have been numerous gifts of specimens for conversion into com- 
parative skeletons; those who provided material used in this study are Don 
B. Alford, Charles M. Bogert, William K. Davis, William G. Degenhardt, 
Kenneth N. Geluso, Dane G. Johnson, Craig B. Jones, Kirkland L. Jones, 
Leonard M. Urban, and Michael A. Williamson. Finally, for the use of 
archeological herpetofauna recovered in the course of her dissertation, I am 
indebted to Susan M. Riches; her original field work was partly supported 
by Grant No. 2391 from the Wenner-Gren Foundation for Anthro- 
pological Research. 



The Biogeographical Relationships of 
the Amphibians and Reptiles of the 
Guadalupe Mountains 



JOHN S. MECHAM, Texas Tech University, Lubbock 

The Guadalupe Mountains of southern New Mexico and adjacent Texas 
possess an unusually rich and diverse herpetofauna. This is undoubtedly due 
in large part to the unusual diversity of habitats within the region, but it may 
also reflect the strategic location of the mountains with respect to biogeo- 
graphic regions. The Guadalupes are essentially an extension of the more 
prominent Sacramento Mountains that lie immediately to the northwest, 
and as such are subject to some biotic influence of the southern Rocky 
Mountain system. However, the southern Great Plains lie to the east of the 
mountains, and the great Chihuahuan Desert stretches away to the south. 
Also, the Pecos River is located only a few miles to the east, and could act as 
an avenue for the intrusion of some more aquatic species. 

The purposes of this paper are to provide a brief summary of the 
amphibians and reptiles that occur in the Guadalupe Mountains and their 
immediate vicinity and to analyze the faunal elements represented in the 
herpetofauna and their relative importance. Needless to say, any con- 
clusions relative to the biogeographic position of the area that are based on 
the herpetofauna alone may not apply to other groups. 

THE HERPETOFAUNA 

Bailey (1905) briefly described the vertebrate fauna of the Guadalupe 
Mountains and wrote (1928) a popular account of the animals found in the 
vicinity of Carlsbad Caverns. The first serious study of the amphibians and 
reptiles as such, however, was that of Mosauer (1932), who recorded 18 
species from the area. My own field work in the Guadalupe Mountains 
resulted in the collection of 50 species, together with considerable relevant 
ecological data (Mecham 1955). Most of this information has not been pub- 
lished, although two taxonomic papers (Mecham 1956, 1957) were stimu- 
lated by this new material. Frederick Gehlbach has done extensive work on 
the ecology and distribution of the herpetofauna, but again much of this 
work remains unpublished. A mimeographed summary (Gehlbach 1964) of 

169 



170 MECHAM 

the amphibians and reptiles of Carlsbad Caverns National Park, however, is 
available together with two relevant taxonomic papers (Gehlbach 1974; 
Gehlbach and McCoy 1965) and the contribution in this volume. Many 
other herpetological records are scattered through the literature. A recent 
review of the turtles of New Mexico (Degenhardt and Christiansen 1974) 
provides a useful summary of the distribution of this group in the region. 
Several studies on late and post-Pleistocene herpetological remains are also 
of some relevance. These include Holman (1970), Gehlbach and Holman 
(1974), and Applegarth (1977). 

A list of the amphibians and reptiles known from the Guadalupe Moun- 
tains or their immediate vicinity is given below, together with brief sum- 
maries of their local distribution. Life belts in the sense of Dice (1943) 
have been used in a loosely descriptive sense, although it is recognized that 
an altitudinally based classification of habitats in the Guadalupe Moun- 
tains is unsatisfactory in some respects. The plains life belt as used here refers 
to habitats below roughly 4200 ft on the lower slopes and the adjacent desert 
plains. The roughlands belt encompasses a wide range of habitats in the 
mountains proper from roughly 4200 ft to 7000 ft or more, and including 
evergreen woodland (sensu Gehlbach 1967). The montane belt includes 
coniferous forest as may occur on the peneplane above approximately 7200 
ft. 

Subspecific names have not been used in this account except where rele- 
vant. Forms rare in the area (known from only one or two records) are indi- 
cated by "R." Forms indicated by an asterisk presumably do not occur 
within the boundaries of the Guadalupe Mountains National Park; other 
species listed have been reported from the park or almost certainly occur 
there. Distributional information is based primarily on Mecham (1955) 
together with published sources except as otherwise noted. 

Amphibia. Recorded species include Ambystoma tigrinum, Scaphiopus 
couchi, Scaphiopus hammondi, Scaphiopus bombifrons, * Hyiactophryne 
augusti (R), Bufo cognatus (R), Bufo debilis, Bufo punctatus, Bufo 
speciosus*Acris crepitans, and Rana berlandieri. The bullfrog, Rana 
catesbeiana, also is present in the region, but apparently is an introduction. 
Ambystoma tigrinum may occur at all altitudes in ponds or tanks. The 
Scaphiopus species, Bufo cognatus, B. speciosus, and B. debilis occur pri- 
marily in the plains life belt below 4800 ft, whereas the more rock-loving B. 
punctatus ranges from the lower altitudes to above 6000 ft. Acris crepitans is 
known only from permanent water sites on the eastern side of the range, as 
are southern leopard frogs {Rana berlandieri). Bufo cognatus is known from 
a single record south of Dell City on the southwestern side of the mountains 
(data of Gehlbach). The robber frog, Hyiactophryne augusti, is known from 
a single specimen taken northwest of Carlsbad (Koster 1946). This was the 
only record of the species in New Mexico for a number of years, although 
another specimen has since been taken near Roswell (Zweifel 1967). 



HERPETOFAUNA 171 

Testudinata. Turtles recorded include *Chelydra serpentina, Kinosternon 
flavescens, *Chrysemys picta, * Chrysemys scripta, *Chrysemys concinna, 
Terrapene ornata, and * Trionyx spiniferus. With the exception of Kino- 
sternon flavescens, which is distributed widely in association with perma- 
nent and semipermanent ponds and streams, the aquatic turtle species are 
essentially limited in the area to the Pecos River and to its tributary, the 
Black River (Degenhardt and Christiansen 1974), which extends to the 
eastern foothills of the mountains. The only terrestrial species, Terrapene 
ornata, is widely distributed at lower elevations. 

Lacertilia. Species recorded include Eumeces obsoletus, Eumeces multi- 
virgatus, Cnemidophorus exsanguis, Cnemidophorus gularis, 
Cnemidophorus inornatus, Cnemidophorus tesselatus, Cnemidophorus 
tigris, Coleonyx brevis, Crotaphytus collaris, * Crotaphytus wislizeni, 
Cophosaurus texanus, Holbrookia maculata (R), Phrynosoma cornutum, 
Phrynosoma douglassi, Phrynosoma modestum, Sceloporus poinsetti, 
Sceloporus undulatus, Urosaurus ornatus, and Uta stansburiana. 

Holbrookia maculata is reported here in the area for the first time. Speci- 
mens in the Texas Tech Museum collection were taken by Mr. Tony Burgess 
from gypsum dunes on the southwestern side on the mountains near Eclipse 
Well. Cnemidophorus tigris, C. inornatus, C. gularis, Crotaphytus wislizeni, 
and Uta stansburiana are all essentially confined to the desert plains below 
4500 ft. Cnemidophorus tigris and Crotaphytus wislizeni apparently have 
been collected in the immediate vicinity only from mesquite dunes border- 
ing the salt flats to the southwest. Forms such as Coleonyx brevis and 
Phyrnosoma modestum occur in parts of the desert plains but also pene- 
trate lower parts of the roughlands belt. Phrynosoma cornutum not only 
occurs widely in the plains belt, but ranges to nearly 6000 ft. Eumeces 
obsoletus, Cnemidophorus tesselatus, Cnemidophorus exsanguis, Cro- 
taphytus collaris, and the rocky adapted Cophosaurus texanus are all 
common in more open roughlands habitats to approximately 6000 ft. The 
saxicolous Urosaurus ornatus and Sceloporus poinsetti have extremely wide 
altitudinal ranges, and S. poinsetti, at least, ranges above 8000 ft. The 
ubiquitous Sceloporus undulatus occurs at all altitudes as does Eumeces 
multivirgatus. Phrynosoma douglassi occurs in evergreen woodland and 
coniferous forest, usually above 6000 ft. The last four species listed all occur 
in the coniferous forest of The Bowl. 

Serpentes. Species recorded include Leptotyphlops dulcis,*Thamnophis 
marcianus,*Thamnophis proximus, Thamnophis cyrtopsis, * Matrix 
erythrogaster, Arizona elegans, Elaphe guttata, Elaphe subocularis, 
Pituophis melanoleucus, Rhinocheilus lecontei, Salvadora grahamiae, 
Salvadora hexalepis (R), Sonoraepiscopa, Diadophis punctatus, Gyalopion 
canum, Lampropeltis getulus, Lampropeltis mexicana (R), Mastic ophis 
flagellum, Masticophis taeniatus, Opheodrys vernalis (?), Heterodon 



172 MECHAM 

nasicus, Hypsiglena torquata, Tantilla atriceps, Tantilla nigriceps, Crotalus 
atrox, Crotalus lepidus, Crotalus molossus, Crotalus scutulatus (R), and 
Crotalus viridis. 

The inclusion of Opheodrys vernalis is based primarily on a recent sight 
record of the species in the McKittrick Canyon area by Mr. Tony Burgess, 
although a rancher some years ago gave the writer a good description of 
what apparently was this species in the vicinity of the ruins of Queen, New 
Mexico (northern Guadalupe Mountains, 6000 ft). The form is known as a 
sub-Recent fossil (Logan and Black 1977), and occurs nearby in the Sacra- 
mento Mountains. The presence of Lampropeltis mexicana is based on a 
single specimen from the vicinity of Pine Springs (Gehlbach and McCoy 
1965), the northernmost record of the species. Inclusion of Salvadora 
hexalepis is based on a specimen in the collection of the Carlsbad Caverns 
National Park (data of Gehlbach). The listing of only one form of Diadophis 
is an over simplification. Gehlbach (1974) found evidence to indicate that 
small (D. p. arnyi) and large (D. p. regalis) forms both occur in the Guada- 
lupe Mountains where they may act as distinct species. The two forms inter- 
grade extensively in other areas of contact, however. 

Distributional patterns of the snakes are complex. Thamnophis 
proximus, Thamnophis marcianus, Thamnophis cyrtopsis, and Matrix 
erythrogaster are all confined to the vicinity of permanent water at lower 
altitudes and drainages on the eastern or northeastern side of the moun- 
tains. Thamnophis cyrtopsis, a form that is less dependent on permanent 
water, is more characteristic of the mountains proper and is of wide 
occurrence. A few species are most characteristic of the plains belt and 
appear to invade the mountains only at lower altitudes. These include 
Arizona elegans, Rhinocheilus lecontei, Heterodon nasicus, Lampropeltis 
getulus, Tantilla nigriceps, Crotalus viridis, and Crotalus scutulatus. The 
last form is marginal in the area. The closest record is a specimen taken by 
Tony Burgess just southwest of the mountains near Eclipse Well. Two saxi- 
colous rattlesnakes (Crotalus lepidus and Crotalus molossus) are common 
in the roughlands belt and extend to the highest altitudes. I have found them 
as high as 7400 and 8200 ft, respectively. Most of the other species of snakes 
appear to be distributed at lower to intermediate altitudes, spanning parts of 
both the roughland and plains belts. The bullsnake, Pituophis melanoleu- 
cus, apparently has the widest ecological tolerance of any of the snakes. I 
recorded one specimen as low as 3600 ft (in mesquite dunes); Mosauer 
(1932) found a specimen at about 8000 ft in pine-fir forest. 

Other Species. Some additional species as yet unreported from the 
Guadalupe Mountains may occur there. Possibilities include Bufo wood- 
housei, Rana blairi, Leptotyphlops humilis, Coluber constrictor, Lampro- 
peltis triangulum, Thamnophis sirtalis, Trimorphodon biscutatus, and 
Sistrurus catenatus, among others. The proximity of records for Bufo wood- 
housei strongly suggest that this species does occur in the area, at least in the 



HERPETOFAUNA 173 

vicinity of the Pecos River. The presence of Coluber constrictor also is par- 
ticularly likely. The form occurs nearby in the Sacramento Mountains and 
has recently been reported to the south in the Davis Mountains (Glidewell 
1974). Stebbins (1951) indicated on a map that the canyon treefrog, Hyla 
arenicolor, is present in the Guadalupe Mountains, but this almost certainly 
was in error. Absence of the species, however, is puzzling in view of the 
seemingly optimal habitat that is present at several locations, particularly in 
McKittrick Canyon. 

BIOGEOGRAPHICAL RELATIONSHIPS 

Dice (1943) and Blair (1950) have placed the upper parts of the Guada- 
lupe Mountains within the Navahonian biotic province, of which they would 
form the southernmost extension. Lower areas were placed within the 
Chihuahuan biotic province. The Navahonian, as identified by Dice, is an 
extensive region that lies between the south-central Rocky Mountains 
(Coloradan biotic province) and the southwestern deserts. In a sense it is a 
zone of transition, and inclusion of much of the Guadalupe Mountains in 
this province emphasizes the northern or montane aspects of the biota. The 
Chihuahuan biotic province corresponds essentially to the Chihuahuan 
Desert, although it is more extensive than the desert proper as identified by 
Shelford (1963) and some other ecologists. 

If the biotic province concept is accepted as a viable method of biogeo- 
graphical classification, this characterization of the Guadalupe Mountains is 
definitely misleading as far as amphibians and reptiles are concerned. This is 
demonstrated by an analysis of the reported forms with respect to their 
occurrence in the Navahonian and Chihuahuan provinces, together with two 
other nearby provinces, the Kansan (corresponding essentially to the 
southern Great Plains) and Balconian (the Edwards Plateau of Texas). As 
shown by Table 1, the strongest affinities lie with the Chihuahuan, with 
somewhat lower but nevertheless strong affinities with the Kansan and 
Balconian. The lowest relationship is with the Navahonian, a finding that is 
highly inconsistent with the classifications of Dice and Blair. Even if the 
analysis is restricted to species that penetrate to higher altitudes, say above 
6000 ft, the herpetofauna appears to be more Chihuahuan than Navahonian. 
At least 18 species occur at such altitudes. These include Ambystoma 
tigrinum, Bufo punctatus, Cnemidophorus exanguis, Eumeces 
multivirgatus, E. obsoletus, Phrynosoma douglassi, Sceloporuspoinsetti, S. 
undulatus, Urosaurus ornatus, Diadophis punctatus, Hypsiglena torquata, 
Masticophis taeniatus, Pituophis melanoleucus, Salvador a grahamiae, 
Thamnophis cyrtopsis, Opheodrys vernalis, Crotalus lepidus, and C. 
molossus. Of these, 15 are important in the Chihuahuan, 13 in the 
Navahonian, 12 in the Balconian, and 7 in the Kansan provinces. 

A somewhat different picture emerges if the amphibians and turtles, a 
greater proportion of which have higher water requirements, are con- 
sidered separately from lizards and snakes (Tables 2 and 3, respectively). It 



174 MECHAM 



TABLE 1. Distribution of amphibians and 
reptiles of the Guadalupe Mountains in near 
biotic provinces. 



Biotic 
provinces 



Important Marginal Total 



Kansan 


40 


4 


44 


Balconian 


41 


7 


48 


Navahonian 


31 


10 


41 


Chihuahuan 


46 


13 


59 



TABLE 2. Distribution of amphibians and 
turtles of the Guadalupe Mountains in near 
biotic provinces. 



Important Marginal Total 
provinces 



Kansan 


14 





14 


Balconian 


14 


2 


16 


Navahonian 


6 


3 


9 


Chihuahuan 


8 


6 


14 



TABLE 3. Distribution of lizards and snakes 
of the Guadalupe Mountains in near biotic 
provinces. 



Biotic 
provinces 



Important Marginal Total 



Kansan 


26 


4 


30 


Balconian 


27 


5 


32 


Navahonian 


25 


7 


32 


Chihuahuan 


38 


7 


45 



may be seen that affinities of the amphibians and turtles are primarily with 
the Kansan and Balconian provinces, although the importance of the Chi- 
huahuan province still ranks above that of the Navahonian. With respect to 
the snakes and lizards, the affinities are overwhelmingly with the 
Chihuahuan province, as would be expected in view of the xeric adaptations 
of many of these forms. 



HERPETOFAUNA 



175 



Another method of analysis is in terms of faunal elements, an approach 
that involves some arbitrary aspects but does provide a different perspec- 
tive. At least seven faunal elements (or categories) appear to be represented 
in the Guadalupe Mountain herpetofauna. These elements (Fig. 1) are listed 
below. 




Fig. 1. Relative importance of the herpetological faunal elements represented in 
the Guadalupe Mountains. Refer to text for further explanation. 



Species with Wide Distributions in the Southwest (28 species, 42.4%).— 
Forms included here are all of such wide distribution that they cannot be 
assigned to any particular biotic or physiographic region. The greater num- 
ber occur in the Chihuahuan Desert, but none have their centers of distribu- 
tion there. Most are limited to arid or semiarid regions in Mexico and the 
western United States, but some ( as Lampropeltis getulus or Ambystoma 
tigrinum) range widely in more mesic regions. 

Chihuahuan Desert Element (12 forms, 18.2%). — This category consists of 
forms that are associated with the Chihuahuan Desert, although they may 
have limited distributions in adjacent areas. Included are Cnemidophorus 
inornatus, C. tigris marmoratus, Coleonyx brevis, Cophosaurus texanus, 



176 MECHAM 

Phrynosoma modestum, Sceloporus poinsetti, Elaphe subocularis, 
Gyalopion canum, Lampropeltis mexicana, Tantilla atriceps, Crotalus 
lepidus, and C. scutulatus. Although Cnemidophorus tigris has a much 
wider distribution, the largely disjunct eastern element of the species 
(primarily C. t. marmoratus) is closely tied to the Chihuahuan Desert and 
may be used as a biogeographic indicator of that region. 
Plains Element (8 forms, 12.1%). — This element is composed of forms that 
have their distributional centers within the Kansanbiotic province (southern 
Great Plains). Included are Scaphiopus bombifrons, Kinosternon 
flavescens, Terrapene ornata, Ewneces obsoletus, Elaphe guttata emoryi, 
Heterodon nasicus, Sonora episcopa, and Tantilla nigriceps. 
Eastern Riparian (Pecos River) Element (eight species, 12.1%). — This dis- 
tinctive element consists of the aquatic and semiaquatic species of mostly 
eastern affinities that have been able to invade (or persist in) the area because 
of the proximity of the Pecos River. All of the forms are associated with 
permanent water at lower altitudes on the eastern or northeastern side of the 
mountains. Included here are Rana berlandieri, Acris crepitans, Chelydra 
serpentina, Chrysemys scripta, C. concinna, Trionyx spiniferus, Natrix 
erythrogaster, and Thamnophis proximus. 

Northern Element (four species, 6.1%). — This small but heterogeneous 
group includes forms that approach or attain their southern limits of distri- 
bution in western Texas in the Guadalupe Mountains region. Two 
yOpheodrys vernalis, Chrysemys picta) are essentially northern and eastern 
forms that occur as post-Pleistocene isolates in the Southwest. Populations 
of the short-horned lizard Phrynosoma douglassi, in both the Guadalupe 
Mountains and the Davis Mountains are disjunct from main elements of the 
species to the north in New Mexico, and also qualify as relicts. Also 
included here is the many-lined skink, Eumeces multivirgatus. This species 
is known to occur further to the south in the Chihuahuan Desert, but is 
primarily associated with the western margin of the Great Plains and the 
Navahonian biotic province. The subspecies (E. m. gaigei) that is found in 
the Guadalupe Mountains has a distribution that corresponds closely to the 
Navahonian. The striped phase of this form predominates at higher alti- 
tudes, whereas the patternless ("taylorF) phase is more common at lower 
altitudes (Mecham 1957). 

"Texan" Element (four species, 6.1%). — This rather artificial category 
includes species (Hylactophryne augusti, Bufo speciosus, Cnemidophorus 
gularis, and Leptotyphlops dulcis) that are primarily limited to Texas in the 
United States, although three have extensive ranges in Mexico. New Mex- 
ico populations of Hylactophryne apparently are isolated from Balconian 
elements of the species, presumably as a result of a climatic trend toward 
increased aridity. 

Other species (two forms, 3.0%). — Two parthenogenetic whiptails, 
Cnemidophorus tesselatus and C. exanguis, do not fit any of the foregoing 



HERPETOFAUNA 177 

categories. Both are about equally distributed between parts of the Chihua- 
huan and Navahonian biotic provinces. 

By way of summary of the various faunal elements, it may be said that 
aside from the large group of species with wide distributions in the South- 
west, the most important contribution is made by the Chihuahuan Desert 
element, with southern Great Plains and eastern riparian elements follow- 
ing in importance. In addition, there is a small but distinctive northern con- 
tribution, and a few forms of importance in Texas-Mexico also reach the 
area. This picture is not inconsistent with that obtained in terms of biotic 
provinces. The Chihuahuan Desert contribution was found to be of out- 
standing importance in both cases, and the northern or Navahonian contri- 
bution was found to be relatively minor. The low importance of the latter is 
not unexpected in view of the decreased importance of terrestrial poikilo- 
therms to the north and in the Rocky Mountains, and should not be viewed 
as indicative of the composition of the biota as a whole. The plains (or 
Kansan) affinities were identified as important in both analyses, and both 
indicated some influence from Texas, although the contribution of the 
Balconian in the biotic province analysis appears to have been exaggerated 
by failure to exclude either widely distributed forms or Pecos River ele- 
ments. 

The overall picture that emerges is the extreme diversity of the herpeto- 
fauna in terms of its origins, and the critical location of the Guadalupe 
Mountains at the contact zone between various biotic or herpetofaunistic 
regions. This is emphasized by the rather remarkable fact that at least 23 of 
the 66 naturally occurring species reported from the Guadalupe Mountain 
area reach their limits of distribution in or near there. This includes eight 
forms of the Chihuahuan Desert, four eastern riparian species, two "Texas" 
species, two northern forms, and seven other southwestern species. 

Some correlations can be seen between the various faunal elements and 
local distributional patterns. This is most obvious, of course, for the Pecos 
River element, members of which are confined to permanent water which is 
found mostly at lower altitudes on the eastern side of the escarpment. Most 
species of the plains element, with the exception of Eumeces obsoletus and 
possibly Kinosternon flavescens, are limited to the plains belt and lower- 
most part of the roughlands belt. The same is essentially true of the Texan 
element. Of the northern forms, both Phrynosoma douglassi and Opheodrys 
vernalis appear to be restricted to higher and more mesic environments. 
Although Eumeces multivirgatus is found at lower altitudes, it is common in 
the highest parts of the mountains. Local distribution of the Chihuahuan 
and other species show no consistent patterns. It is interesting to note in this 
connection that Crotalus lepidus and Sceloporus poinsetti, two Chihua- 
huan Desert species that reach their northern limits of distribution in 
southern New Mexico, together with Crotalus molossus, which ranges only 
slightly further north, all occur at the highest altitudes in the montane belt. 



178 MECHAM 

This striking inconsistency between altitudinal range and northern distribu- 
tional limits may be the result of continual local dispersal of these sax- 
icolous reptiles from lower habitats. Without this replenishment, the high 
altitude populations might not be able to maintain themselves over a long 
period of time. 

Finally, attention should be drawn to the fact that there are considerable 
differences between the herpetofaunas at lower altitudes on the eastern and 
western sides of the mountains. As already emphasized, the eastern riparian 
elements essentially are limited to the eastern side, primarily because of 
proximity of the Pecos River and associated drainage patterns, but also 
because of the scarcity of permanent water on the western face. Thamnophis 
marcianus, although not included in this element, also conforms to this local 
pattern. On the other hand, extreme desert habitats including sandy or dune 
situations near the salt flats on the southwestern side support certain species 
(Crotalus scutulatus, Crotaphytus wislizeni, Cnemidophorus tigris, 
Holbrookia maculata) that apparently are absent on the immediate eastern 
side. Limited information suggests that local distribution of some other 
species is consistent with this dichotomy, but more documentation is needed. 

LITERATURE CITED 

Applegarth, J. S. 1977. Environmental implications of herpetofaunal remains 
from archeological sites west of Carlsbad, New Mexico. This volume. 

Bailey, V. 1905. Biological survey of Texas. TV. Am. Fauna 25:1-222. 

1928. Animal Life of the Carlsbad Cavern. Williams & Wilkins Co., Balti- 
more, 195 pp. 

Blair, W. F. 1950. The biotic provinces of Texas. Tex. J. Sci. 2:93-117. 

Degenhardt, W. G., and J. L. Christiansen. 1974. Distribution and habitats of 
turtles in New Mexico. Southwest. Nat. 19:21-46. 

Dice, L. R. 1943. The Biotic Provinces of North America. Univ. Michigan Press, 
Ann Arbor, 78 pp. 

Gehlbach, F. R. 1964. Amphibians and reptiles of Carlsbad Caverns National 
Park, New Mexico, and adjacent Guadalupe Mountains, Mimeographed, 12 pp. 

1967. Vegetation of the Guadalupe Escarpment, New Mexico-Texas. 

Ecology 48:404-419. 

1974. Evolutionary relations of southwestern ringneck snakes {Diadophis 



punctatus). Herpetologica 30:140 148. 

Gehlbach, F. R., and J. A. Holman. 1974. Paleoecology of amphibians and 
reptiles from Pratt Cave, Guadalupe Mountains National Park, Texas. South- 
west. Nat. 19:191-197. 

Gehlbach, F. R., and C.J. McCoy. Jr. 1965. Additional observations on 
variation and distribution of the gray-banded kingsnake, Lampropeltis mexi- 
cana (Garman). Herpetologica 21:35-38. 

Glidewell, J. 1974. Records of the snake Coluber constrictor (Reptilia: 
Colubridae) from New Mexico and the Chihuahuan Desert of Texas. Southwest. 
Nat. 19:215-217. 

Holman, J. A. 1970. A Pleistocene herpetofauna from Eddy County, New Mex- 
ico. Tex. J. Sci. 22:29-39. 



HERPETOFAUNA 179 

Koster, W. J. 1946. The robber frog in New Mexico. Copeia 1946:173. 

Logan, L. E., and C. C. Black. 1977. The quaternary vertebrate fauna of Upper 

Sloth Cave, Guadalupe Mountains National Park, Texas. This volume. 
Mecham, J. S. 1955. A biogeographic study of the reptiles and amphibians of the 

Guadalupe Mountains. Ph.D. thesis, Univ. Texas, Austin, 118 pp. 
1956. The relationship between the ringneck snakes Diadophis regalis and 

D. punctatus. Copeia 1956:51-52. 
1957. The taxonomic status of some southwestern skinks of the tnulti- 



virgatus group. Copeia 1957: 1 1 1-123. 
Mosauer, W. 1932. The amphibians and reptiles of the Guadalupe Mountains of 

New Mexico and Texas. Occas. Pap. Mus. Zool. Univ. Mich. 246:1-18. 
Shelford, V. E. 1963. The Ecology of North America. Univ. Illinois Press, Ur- 

bana, 610 pp. 
Stebbins, R. C. 1951. Amphibians of Western North America. Univ. California 

Press, Berkeley and Los Angeles. 
Zweifel, R. G. 1967. Eleutherodactylus augusti. Catalog. Am. Amphib. Reptiles 

41:1-41.4. 



ACKNOWLEDGMENTS 

Long delayed thanks are expressed for the help of Drs. James R. Tamsitt, 
M. Jack Fouquette Jr., and Robert H. DeWolfe, all of whom accompanied 
the writer on collecting trips during our student days. The writer would also 
like to express his appreciation to authorities of the Carlsbad Caverns 
National Park for the use of the park collection, to Mr. Tony Burgess of 
Texas Tech University, who has provided several new distributional records, 
and to Dr. Frederick R. Gehlbach, who kindly made available his unpub- 
lished summary of locality records of museum specimens from the Guada- 
lupe Mountains. 



Compositional Aspects of Breeding 
Avifaunas in Selected Woodlands 
of the Southern Guadalupe 
Mountains, Texas 



GEORGE A. NEWMAN, Hardin-Simmons University, 
Abilene, Texas 

The Guadalupe Mountains, located in the Trans-Pecos area of west 
Texas and in the southeast portion of New Mexico, have been the object of 
numerous geological investigations (e.g., Cys, 1971; Newell et al. 1972). 
Perhaps this mountain range is best known for its unique Carlsbad Caverns 
system, situated at the northern end of the mountain mass. At the southern 
end of the Guadalupe complex, however, are found landmarks which were 
of great importance to man long before the discovery of the now famous 
caverns. El Capitan with its uninterrupted cliff facing rising more than 2000 
ft (610 m) and Guadalupe Peak, the highest peak in Texas at 8791 ft (2679.5 
m), have provided massive landmarks utilized by early explorers and 
travelers (Levy 1971). 

The federal government recognized the scientific, aesthetic, and historical 
value of the southern Guadalupe Mountains by creating Guadalupe 
Mountains National Park on 15 October 1966 (Public Law 89-667). This 
relatively new national park consists of 77,500 acres (31,364.3 ha). Its 
northern boundary is the New Mexico-Texas state line and the southern 
boundary encompasses El Capitan and Guadalupe Peak. An assemblage of 
diverse habitats ranging from Chihuahuan Desert, surrounding the wedge- 
shaped mountain mass, to coniferous forest at the high mountain elevations 
characterize this unique area. Rugged canyons penetrate deep into the 
interior of the mountains, providing environmental conditions which 
support mesic relict woodland communities (Gehlbach 1967). 

Published accounts of bird investigations pertaining to the Guadalupe 
Mountains date back to 1901 to the work of Mr. and Mrs. Vernon Bailey 
(Bailey 1905). Burleigh and Lowery (1940) reported on their findings result- 
ing from several collecting excursions to the Guadalupes in 1938 and 1939. 
At about the same time as Burleigh and Lowery, W. B. Davis, along with 

181 



182 



NEWMAN 



several of his students, made collections and recorded bird observations 
from the southern Guadalupe Mountains (Biaggi 1960). However, these 
early accounts provided little quantitative information on breeding birds of 
the Guadalupes. 

Investigations of bird populations of other southwestern mountain ranges 
have been accomplished by Marshall (1957) in southern Arizona, Hubbard 
(1965) in the Mogollon Mountains of New Mexico, Tatschl (1967) in the 
Sandia Mountains of New Mexico, Wauer (1971) in the Chisos Mountains 
of Texas, Carothers et al. (1973) in the San Francisco and White Mountains 
of Arizona, and Johnson (1974) in the Grapevine and Potosi mountains of 
Nevada. 

In the summers of 1969, 1970, and 1971, at the invitation of the National 
Park Service, I began preliminary investigations of the breeding avifaunas of 
selected woodland communities within Guadalupe Mountains National 
Park (Newman 1971, 1974a,b). Beginning in 1972 and continuing through 



\ ^>) 



Q 



8000 



A % %) ** } §j ) 

|0 flP> 0^5fc,» oO °" 7*°° 



\ ■ 



r. 



N 



w j} 



-ssoo- 







Contour Intervol 500 Feci 



FIG. 1. — Map of southern Guadalupe Mountain region, adapted from the U.S. 
Geological Survey map of the Guadalupe Peak Quadrangle. Inset shows the relation 
of the southern Guadalupe Mountains to the remainder of the state of Texas. 



BREEDING AVIFAUNAS 183 

the summers of 1973 and 1974, 1 made intensive surveys of the breeding birds 
in Main McKittrick Canyon, South McKittrick Canyon, North McKittrick 
Canyon, Upper Dog Canyon, and The Bowl. 

The purposes of this study were: (1) to quantitatively document the 
breeding avifaunas of relict woodlands of the Southern Guadalupe 
Mountains; (2) to determine the stability of the breeding avifaunas on a year 
to year basis; (3) to compare selected breeding avifaunas of the southern 
Guadalupe Mountains with those previously studied by other scientists in 
other southwestern mountain ranges. 



METHODS 

Five study plots were established in the summer of 1972, one each in the 
following areas: Main McKittrick Canyon; North McKittrick Canyon; 
South McKittrick Canyon; Upper Dog Canyon; The Bowl (Fig. 1). In the 
canyon areas, study plots were established according to the natural terrain 
and were limited to the canyon floors and lower terraces. The study plot in 
The Bowl was established in an area showing little disturbance by man. 
Study plot sizes were constant at 24 acres (9.7 ha) for South McKittrick Can- 
yon, Upper Dog Canyon, and The Bowl. The Main McKittrick and North 
McKittrick study plots consisted of 30 acres ( 12. 1 ha) and 20 acres (8. 1 ha), 
respectively. The variation in size in the Main McKittrick and North 
McKittrick plots was due to natural terrain factors which dictated plot 
boundaries. Plot sizes were determined with the aid of a planimeter and 
aerial photographs. All boundaries of study plots were identified according 
to natural features and were constant throughout the 3-year study period. 

Expeditions to the Guadalupes lasted from late May through June each 
summer, 1972 to 1974. Plots were visited in the same rotation each 
summer — McKittrick Canyon system, then Upper Dog Canyon, and finally 
The Bowl. All three McKittrick study plots were easily accessible from a 
camp at Pratt Lodge and time was divided equally among these three areas. 
Eight to 10 days each summer were spent working Upper Dog Canyon from 
a camp at the Ranger Station. The Bowl was worked during the latter part of 
June each summer for a period of 6 to 10 days from a primitive camp. 

Vegetation descriptions of the study areas were based on tree data col- 
lected by the author, during the summer of 1974, using the Cottam and 
Curtis (1956) point-quarter method. Only those trees with a minimum 
circumference of 30 cm were included in the sampling procedure. An 
analysis of the vegetation of the Guadalupe escarpment has been presented 
by Gehlbach (1967). 

Techniques adapted from Kendeigh (1944) were used to measure the 
absolute numbers of breeding birds per study plot. A combination of early 
morning censuses, searches for nests, and limited mist-netting were carried 
out on each study plot. The steep canyon walls set well-defined limits on can- 
yon plots and so it was possible to determine accurately what birds were truly 



184 



NEWMAN 



breeding within them. Singing-male censuses were taken by slowly walking 
through the study plots, with periodic pauses to record singing males on 
specially designed census cards. All singing-male censuses began at sunrise 
(ca. 06:00 MDST) and lasted about 2 hours (08:00 MDST). Other daylight 
hours were utilized in searching for nests and in netting activities. The 
absolute numbers of breeding birds per study area were expanded to the 
standardized expression, numbers of birds per 100 acres. Observations were 
augmented by the use of Leitz Trinovid 10 * 40 binoculars. 

Of more ecological meaning than numbers of birds per unit area is grams 
of biomass per unit area. Biomass computations for breeding birds were 
based on average fresh weights of specimens collected from near the study 
plots over a 6-year period (1969-1974), except for the weights of the Elf Owl 
(Micrathene whitneyi) and the Blue-throated Hummingbird (Lampornis 
clemenciae) which were taken from specimens in the Texas Cooperative 
Wildlife Collections of Texas A&M University. Standing crop biomass was 
computed by multiplying the density of a species by the mean weight of that 
species. 

Consuming biomass, an indication of food consumption, was figured as 
suggested by Salt (1957). The average weight of each species was raised to the 
0.633 power (Karr 1968) and then multiplied by the number of individuals 
per species present per 100 acres (40.47 ha). The summation of the values 
thus obtained for each species gives an indication of energy requirement for a 
given avifauna. 

Species were assigned to foraging categories in ways similar to those used 
by Salt (1953, 1957) and Anderson (1970). Assignment to categories was on 
the basis of field observations made during the breeding seasons and on the 
basis of published data on feeding habits of the species under consideration. 
Where a species fitted into more than one foraging group, it was placed in the 
group which characterized more than 50 percent of its feeding effort. 
Categories used in this study were foliage insect, aerial insect, ground insect, 
timber drilling, timber searching, foliage seed, foliage nectar, ground seed, 
and ground predator. 

Species diversity indices of breeding avifaunas were computed by the 
Shannon-Weiner index (//' = -Dpi loge pi). Coefficients of similarity as 
described by Beals (1960) were computed and used to compare the avi- 
faunas of each study area. 

RESULTS 

Main McKittrick Canyon 

Habitat Description — Main McKittrick Canyon cuts deep into the east face 
of the Guadalupe escarpment. The overall direction of drainage in Main 
McKittrick is from west to east, though the course of the stream bed varies. 
From the mouth of Main McKittrick to the junction of North and South 
McKittrick along the canyon floor is a distance of about 1.8 mi. (2.9 km). 



BREEDING AVIFAUNAS 



185 



Canyon walls rise abruptly to a height of 2000 ft (610 m) above the canyon 
floor. The floor of the canyon varies in width but is not much greater than 
360 ft (110 m) at its widest sect and gradually narrows towards the west. 
Elevation of the canyon floor increases from 5000 ft ( 1 524 m) at the mouth of 
the canyon to 5200 ft (1585 m) at the junction of North and South 
McKittrick. The vegetational growth form along the canyon floor gradually 
shifts from shrub and grass domination near the mouth of the canyon to tree 
domination. Above-ground water exists year-round in Main McKittrick. 
The stream flow is intermittent in space, alternately flowing above and 
below ground. This flow is from drainage of precipitation and from active 
springs in North and South McKittrick canyons. 

The east boundary of the Main McKittrick study plot is located about 0.8 
mi. (1.3 km) in from the mouth of the canyon and runs a linear distance of 
about 1.0 mi. (1.6 km). 

One hundred and twenty trees were measured on the study plot. A com- 
parison of the relative density, relative dominance, relative frequency, and 
importance value of trees in Main McKittrick, South McKittrick, North 
McKittrick, Upper Dog, and The Bowl is shown in Tables 1-4. Wavy-leaf 
oak (Quercus undulata), alligator juniper {Juniperus deppeana), and Texas 
madrone (Arbutus texana), were relatively important (Table 4) in Main 
McKittrick with values of 80.17, 67.81, and 52.52, respectively. Common 



TABLE 1. Relative density, obtained by the point-quarter method, of tree species in Main 
McKittrick (1), North McKittrick (2), South McKittrick (3), Upper Dog (4), and The Bowl 
(5). 



Species 




Relative density 






1 


2 


3 


4 


5 


Quercus undulata 


32.50 


16.67 


11.67 






Juniperus deppeana 


20.00 


22.50 


13.33 


9.00 


0.83 


Arbutus texana 


15.83 


12.50 


6.11 


1.00 




Acer grandidentatum 


10.00 


16.67 


27.22 


28.50 




Quercus muhlenbergia 


7.50 


20.00 


25.00 


38.50 




Fraxinus velutina 


5.00 




3.89 






Pinus ponderosa 


3.33 


1.67 


9.44 


7.00 


30.42 


Prunus serotina 


2.50 


1.67 


1.11 




0.83 


Juglans microcarpa 


1.67 










Juniperus sp. 


0.83 




1.11 






Chilopsis linearfolia 


0.83 










Ostrya knowltonii 




4.17 




3.00 




Juniperus scopulorum 




2.50 




2.00 




Pseudotsuga menziesii 




1.67 




3.50 


18.33 


Quercus grisea 






1.11 






Quercus gambeli 








7.00 


17.08 


Juniperus monosperma 








0.50 




Pinus flexilis 










32.50 



186 NEWMAN 



TABLE 2. Relative dominance, obtained by the point-quarter method, of tree species in Main 
McKittrick (1), North McKittrick (2), South McKittrick (3), Upper Dog (4), and The Bowl 
(5). 



Species 



Relative dominance 



1 


2 


3 


4 


5 


23.48 


10.90 


7.71 






26.15 


28.36 


10.98 


12.45 


0.40 


18.64 


12.80 


6.43 


0.78 




7.08 


12.25 


20.97 


22.20 




8.70 


25.26 


31.17 


36.98 




5.44 




7.21 






3.45 


1.74 


12.67 


11.56 


24.30 


2.57 


1.66 


1.11 




0.58 


2.07 










1.64 




1.00 






0.77 


2.37 
3.13 




2.07 
3.09 






1.55 




5.01 


28.89 



Quercus undulata 

Juniper us deppeana 

Arbutus texana 

Acer grandidentatum 

Quercus muhlenbergia 

Fraxinus velutina 

Pinus ponderosa 

Prunus serotina 

Juglans microcarpa 

Juniper us sp. 

Chilopsis linearfolia 

Ostrya knowltonii 

Juniperus scopulorum 

Pseudotsuga menziesii 

Quercus grisea 0.78 

Quercus gambeli 5.54 11.18 

Juniperus monosperma 0.48 

Pin us flexilis 34.66 



TABLE 3. Relative frequency, obtained by the point-quarter method, of tree species in Main 
McKittrick (1), North McKittrick (2), South McKittrick (3), Upper Dog (4), and The Bowl 
(5). 



Species 



Relative frequency 



1 


2 


3 


4 


5 


24.19 


15.56 


11.92 






21.66 


20.62 


12.69 


12.12 


1.28 


18.05 


12.84 


6.92 


1.52 




9.75 


15.56 


27.31 


25.76 




8.30 


19.46 


23.08 


32.58 




6.14 




4.23 






3.61 


2.72 


9.23 


8.33 


25.53 


3.61 


2.72 


1.54 




1.28 


2.53 










1.08 




1.54 






1.08 


3.89 
3.89 




3.79 
3.03 






2.72 




5.30 


20.43 



Quercus undulata 

Juniperus deppeana 

Arbutus texana 

Acer grandidentatum 

Quercus muhlenbergia 

Fraxinus velutina 

Pinus ponderosa 

Prunus serotina 

Juglans microcarpa 

Juniperus sp. 

Chilopsis linearfolia 

Ostrya knowltonii 

Juniperus scopulorum 

Pseudotsuga menziesii 

Quercus grisea 1 .54 

Quercus gambeli 6.82 20.43 

Juniperus monosperma 0.76 

Pinus flexilis 31.06 



BREEDING AVIFAUNAS 



187 



TABLE 4. Importance value, obtained by the point-quarter method, of tree species in Main 
McKittrick (1), North McKittrick (2), South McKittrick (3), Upper Dog (4), and The Bowl 
(5). 



Species 



Importance value 



1 


2 


3 


4 


5 


80.17 


43.13 


31.30 






67.81 


71.48 


37.00 


33.57 


2.51 


52.52 


38.14 


19.46 


3.30 




26.83 


44.48 


75.50 


76.46 




24.50 


64.72 


79.25 


108.06 




16.58 




15.33 






10.39 


6.13 


31.34 


26.89 


80.25 


8.68 


6.05 


3.76 




2.69 


6.27 










3.55 




3.65 






2.68 


10.43 
9.52 




8.86 

8.12 






5.94 


3.43 


13.81 

19.36 

1.74 


67.65 
48.69 
98.22 



Quercus undulata 
Juniper us deppeana 
Arbutus texana 
Acer grandidentatum 
Quercus muhlenbergia 
Fraxinus velutina 
Pinus ponderosa 
Prunus serotina 
Juglans microcarpa 
Juniper us sp. 
Chilopsis linearfolia 
Ostrya knowltonii 
Juniperus scopulorum 
Pseudotsuga menziesii 
Quercus grisea 
Quercus gambeli 
Juniperus monosperma 
Pinus flexilis 





r-v • -$r 



**■*+ 



J* 



FIG. 2. — Main McKittrick Canyon, Guadalupe Mountains National Park, Texas. 



188 NEWMAN 



TABLE 5. Density of breeding bird populations in Main McKittrick Canyon for 1972, 1973, 
and 1974. 



Species 



Pairs 


per 100 


acres 


1972 


1973 


1974 


23.3 


16.7 


23.3 


20.0 


6.7 


16.7 


20.0 


20.0 


13.3 


16.7 


20.0 


16.7 


16.7 


10.0 


23.3 


16.7 


6.7 


13.3 


13.3 


13.3 


10.0 


13.3 


10.0 


13.3 


13.3 


13.3 


10.0 


10.0 


10.0 


6.7 


10.0 


10.0 


10.0 


10.0 


3.3 


6.7 


10.0 


6.7 


10.0 


10.0 


6.7 


3.3 


6.7 


3.3 


6.7 


6.7 


6.7 


6.7 


6.7 


10.0 


13.3 


3.3 


0.0 


6.7 


3.3 


6.7 


0.0 


3.3 


3.3 


0.0 


3.3 


0.0 


6.7 


3.3 


0.0 


0.0 


3.3 


3.3 


3.3 


3.3 


0.0 


3.3 


3.3 


0.0 


0.0 


3.3 


6.7 


10.0 


0.0 


3.3 


6.7 


0.0 


3.3 


3.3 


0.0 


3.3 


0.0 


0.0 


3.3 


0.0 


0.0 


0.0 


3.3 


0.0 


0.0 


3.3 


0.0 


0.0 


3.3 


0.0 


0.0 


3.3 


253.1 


206.6 


256.5 



Blue-gray Gnatcatcher {Polioptila caerulea) 
Bewick's Wren ( Thryomanes bewickii) 
Rufous-crowned Sparrow {Aimophila ruficeps) 
Solitary Vireo ( Vireo solitarius) 
Black-chinned Sparrow {Spizella atrogularis) 
Canyon Wren {Catherpes mexicanus) 
Black-chinned Hummingbird (Archilochus alexandri) 
Ash-throated Flycatcher {Myiarchus cinerascens) 
Western Tanager ( Piranga ludoviciana) 
Cassin's Kingbird {Tyrannus vociferans) 
Western Wood Pewee {Contopus sordidulus) 
Violet-green Swallow ( Tachycineta thalassina) 
Bushtit {Psaltriparus minimus) 
Lesser Goldfinch {Spinus psaltria) 
Scott's Oriole {Icterus parisorum) 
Brown-headed Cowbird {Molothrus ater) 
Black-headed Grosbeak (Pheucticus melanocephalus) 
Broad-tailed Hummingbird {Selaphorus platycercus) 
Ladder-backed Woodpecker {Dendrocopos scalaris) 
Scrub Jay (Aphelocoma coerulescens) 
House Wren {Troglodytes aedori) 
Gray Vireo ( Vireo vicinior) 
Hepatic Tanager {Piranga flava) 
Blue Grosbeak ( Guiraca caerulea) 
House Finch {Carpodacus mexicanus) 
Rufous-sided Towhee {Pipilo erythrophthalmus) 
Virginia's Warbler ( Vermivora virginiae) 
Olive-sided Flycatcher 2 {Nuttallornis borealis) 
Elf Owl {Micrathene whitneyi) 
Brown Towhee {Pipilo fuscus) 
Western Flycatcher {Empidonax difficilis) 
Rock Wren {Salpinctes obsoletus) 
Hermit Thrush {Catharus guttatus) 
Warbling Vireo ( Vireo gilvus) 
Total 



a Breeding status not certain. 



shrubs and grasses contributing to the understory in Main McKittrick were 
Quercus undulata, Juniperus, deppeana, Acer grandidentatum, Dasylirion 
leiophyllum, Juglans microcarpa, Agave neomexicana, Yucca elata, 
Cladium jamaicensis, Bouteloua curtipendula, and Tridens muticus. 
Physiognomic aspects of Main McKittrick Canyon are shown in Fig. 2. 

Avifaunal Aspects. — Densities of the breeding avifaunas of Main McKit- 
trick Canyon for 1972, 1973, and 1974 are presented in Table 5. Thirty-four 



BREEDING AVIFAUNAS 



189 



s§ 



Tj" fN IT) t— 



On fN O CN CO On 



Oiooo^TNO-^-coOOOn 

r-H CN fN r-i r-i 



On ^ oo o> r- vo 

rtV m oo C^ iH 

>/-) ro r- On v~> co 
— i fN 



CO O 

u © 

CO ^ 



OXHNf^OMO 
(JiCOTtHOOfOOOH 

in 10 - i uo cn cn — i -h 



5 «o 


i-H 


00 


lO 


ON 


m 


O 


NO 


l-Tj 


5 tT 


m 


On 


o 


»o 


CO 


o 


fN 


CO 


o 


00 


tj- 


T}- 


Tt 


on 


I/O 


O 


f- 


1-1 




oo 

CO 


\o 


CO 


'" H 


fN 


r\ 





s§ 



NO 


r- 


ON 


m 


o> 


— i o 


co 


o 


CN 


o 


o 


CO 


r» 


rn 


ON 


t-~ 


O 


ON 


NO 
ri 
n 




NO 
C I 


NO 


<* 

o 


Tf O 

NO 


o 


o 


'St 


o 


On 
00 
ON 


CO 
fN 


co 


co 


■CO 

o> 




o 


fN 

o 

NO 






^-oooot^Noo^toioo r- 

t^tri^^o^OHO no 

oo o ^o m o co r~ io r- 

t-» r- oo fN fN fN ^h 



O no o 

d 



>o no r» o <n 

W~) CO ON i/-} O 
On vo fNi cN 



<U NO fN CO r~ Tt- 



5 2 



NO »/-) lO ^t NO 
fN O -^ CO Tt 

(S (N H H rt 



^h ON fN ^h O O 

tj- co d co d d 

NO lO NO CO 



m 


NO 


1-1 


On 


o o 


On 






O 


CO 
ON 


o o 


NO 



00 O 4) 
U © g 



000>fNCONOfNO<0 



O O fN 



^tr-OOfOTj-ONfNrfTtOO 

oo oo ^ - < oo m oo h oo 



O On O O © © On 

no >o no d d d c*- 

m rfr On u-> ■>* 

On 00 fN fN CO 
fN 



inNOfNco— iconoc— ooooon 



oq oq oo io t>- on 
r^ i-5 Tf fN d i-5 

Tfr CO ^H r-l CO -H 



C/3 



O S 



<rs .3 
H > 



<u 


~3 

r 


0) 

to ^ 

PS TO 






43 


o 

> 




O 


O 
>i 

M 


PS X) 

■la 


+J 


O 

u 

> 


"pS 


73 
o 

Jo 

C/3 


o 


C3 




43 

09 

3 


>> 

l-i 


.a 


43 

rS 



3 * 

PS o 

*a >. ps 

• S ° 

* 2 






is 

r-0 


0- 

■a 




o 

PS 






D 


; 


o 










>, 




1) 




•a 
u 




O 






& 




Xi 


_H 




6 

> 


<u 


3 






U <> ^ O =3: 



190 



NEWMAN 



5 2 



MO S 

u 2 8 



m 8 
3 2 



oa o 
u © 



e§ 



CO O 

u © 



3 bb O 



to oo co m o\ o q 
o o 



in CO On tJ- VO 

> fO W 0\ N 



oo Tt m o r~ <o on 

On CO CN O f-H sC CO 
CO CN v£> in CO ^h 



o 


On 


CO 


o 


r~- 


o o 


«*■ 


N£> 


VD 


CO 


O 


CO 


CO 


no 


o 


o 


-t 


P<» 


0. 


vo 


O 


o o 


00 


o 


o 


(N 


o 


o 


00 


CN 


o 


o 


CN 


o 


»-H 


cn 


^O 




CO 


CN 


c 


m 




r— 


ON 


-* 






m 


in 


VO 


r- 






'St 

CN 


CO 


en 


•^ 




'"■ ' 




o 







On OO 


o\ 


o 


NO 


O CO 


t*» 


~H 


CN 


o 


CO 


o 


o 


— < 


r- 


r» 


CO CO 
NO CO 


CI 

O 


o 

ro 


io 


O iO 


o 

NO 


lO 


no 


o 


CN 


o 


o 


oo 


o 


O 



o 


On 


o 


'St 


ro 


O 


On 


wo 


no 


-1- 


O 


r» 


o 


o 


r^ 


CO 


CO 


CO 
CO 


o 

IT) 


NO 

CI 


^O 

CO 

<st 


CO 

CN 


o 


OO 
O 
CO 


o 
oo 
"St 


OO 
CN 


lO 


O 


CO 

CN 


o 


o 


to 


NO 
ON 
Ti- 


NO 

ON 



CD 
CO 


O) 


CO 


CO 


l-H 


in 


"St 


o 


in 


CD 


'St 


O 


CO 


q 


o 


o 


CN 


<D 
CD 

cn 


c- 


p- 


ro 


CO 


,H 


"t 


<-H 


•* 


o 


ON 


c 


r^ 


in 


O 


d 


6 


d 


CO 


^4 


,— J 


C 
O 


SO 


ro 


r» 


NO 


CO 


"t 




lO 


<—> 


vO 


in 


ro 








w> 


GO 


o 


o 


CN 


wH 


l-H 










p- 


T3 


l-H 


r-i 










CO 


-— i 


— < 


















C 

3 
















_cfl 






















O 
















"o 






6 


















<5 
















b 








o 


o^ 


CN 


^O 


o 


o 


o 


c^ 




q 


'St 


n 


q 


q 


o 


lO 




CO 


ro 




CO 


r- 


*t 


ON 


'St 


-r 


o 


r^ 




>st 


r- 


in 


d 


© 


d 


NO 




NO 


SO 




CO 


o 


-+ 


co 


CO 


en 




On 




CO 


r^ 


r- 








CO 




On 


On 




r- 


in 


Tf 


CN 


'~ l 


1—1 




CN 
CN 




ro 


CO 










CO 




"St 


"t 



r~ On CO 


CO CN CO 00 


n£> co Tf <-H oo On 


CN 


On t^ co 

^H CO i-H 


NO On O no 

CO CN Tf 


On ^^ -h uo io rj- 

-H rH CO CN r-H 


lO 



I s 

•« £ -S 



2 u ^ 
:y o £ o 
1 O £ H 



=3 in <" 

Q <D V5 



eu 



o 



i2 3 S o 2 

pq cC J ffi « 



i3 o i (u o 



o 


>. 

cd 


-O 


*~> 


3 


X) 


CO 


P 



(SUlWltt! 



BREEDING AVIFAUNAS 



191 



aa 8 
8 2 



OQ O 

u © 

</3 »H 



(N O 
■^t CO 



8 » 



CO o 
U o 

C/3 "- 1 



o 


© 


O (N 


CN 


OO 


d 


d 


M3 OO 


^ 


d 






VO Tf 


- 


OS 



NO o 


vo 


^ 


so d 
to 


NO 


o< 



co 


CO 


OO o 


OO 


NO 


d 


d 


r»' d 


l> 


lO 


c^ 


t~~ 


OO 


00 


NO 


t 


■t 






CO 
OS 









toO 








G 












s§ 






CN 






x> 








S 


rt 






H 




ca O 


t/3 






U o 


u, 






ca ^ 







if® 



S oo oo 



no 


00 


-* 


o 


no 


Tf 


c- 


no 

On 

o 



r— co 


,— 1 


_ 


00 CN 


^H 


ON 




,—l 


no 

ON 



CO NO 
CO CO 



a 

T3 

o 
o 

T3 
O 03 



192 



NEWMAN 



species were recorded as breeding on the study plot over the 3-year period. 
Nineteen (56%) of the 34 species were represented as breeding birds in all 
three of the summers. Fifteen species (44%) were present as breeders in 1 or 
2 of the 3 years. The greatest number of breeding species for season 

was 28 in 1974, followed by 26 in 1972, and 25 in 1973. The highest density 
of birds occurred in 1974 with 256.5 pairs per 100 acres, followed by 253.1 
pairs in 1972, and 206.6 pairs in 1973. 

Standing crop biomass and consuming biomass of breeding avifaunas, 
expanded to reflect grams per 100 acres, are presented in Table 6. The years 
1972 and 1974 were similar in total avifaunal biomass. In 1973 standing crop 
biomass was about 3% lower than it was in 1972 and 1974. Consuming bio- 
mass in 1973 showed a decrease of about 9% from 1972 and 1974. 



C.B. 

% 



S.C.B. 20 

% 




ilfi 



A — Foliage Insect 
B — Aerial Insect 
C — Ground Seed 
D — Ground Insect 
E — Foliage Seed 
F — Timber Drilling 
G — Foliage Nectar 
H — Timber Searching 
I — Ground Predator 



FIG. 3. — Percentage contribution of foraging classes to standing crop biomass 
(SCB) and consuming biomass (CB) for Main McKittrick Canyon. Bars represent 
1972, 1973, and 1974, respectively. 



Percentage contributions of each foraging category to standing crop bio- 
mass and consuming biomass are shown in Fig. 3. Bird species of the foliage 
insect category accounted for the greatest percentage of biomass each of the 
3 years. 

Species diversity for the breeding avifaunas of Main McKittrick Canyon 
was 3.05 for 1972, 3.04 for 1973., and 3.16 for 1974. 



BREEDING AVIFAUNAS 



193 



North McKittrick Canyon 

Habitat Description.— At the west end of Main McKittrick there is a bi- 
furcation of the canyon into north and south branches. North McKittrick 
Canyon meanders to the north for a linear distance of about 1.5 mi. (2.4 km) 
to the New Mexico-Texas state boundary. The canyon floor is narrower 
than that of Main McKittrick. Above-ground water is limited, except for 
flash flood conditions, to the northern few hundred yards of my study plot. 
This plot extends from near the junction of the 3 canyons to the first cross- 
ing of the state boundary. Elevation of the canyon floor along the North 
McKittrick plot gradually increases from 5200 ft (1585 m) at the south end 
to 5400 ft (1646 m) at the north end. Steep canyon walls rise abruptly from 
the canyon floor. 

Data were collected from a total of 120 trees on this plot. Of the 10 species 
represented in North McKittrick, alligator juniper and chinquapin oak 
(Quercus muhlenbergia) dominated the landscape (Tables 1-4). Importance 
values of 71.48 and 64.72, respectively, were recorded (Table 4). Common 
shrubs contributing to the understory in North McKittrick were Quercus 
undulata, Juniperus deppeana, Acer grandidentatum, Dasylirion 
leiophyllum, and Agave neomexicana. Grass cover is quite limited on the 
North McKittrick plot. 

Figure 4 shows the physiognomy of the vegetation in this canyon. 



> ' 




FIG. 4.— North McKittrick Canyon, Guadalupe Mountains National Park, Texas. 



194 



NEWMAN 



TABLE 7. Density of breeding bird populations in North McKittrick Canyon for 1972, 1973. 
and 1974. 



Species 



Pairs 


per 100 


acres 


1972 


1973 


1974 


40.0 


20.0 


25.0 


35.0 


20.0 


45.0 


30.0 


30.0 


35.0 


30.0 


25.0 


30.0 


30.0 


20.0 


35.0 


30.0 


15.0 


35.0 


25.0 


20.0 


20.0 


25.0 


10.0 


20.0 


20.0 


20.0 


35.0 


20.0 


20.0 


25.0 


20.0 


15.0 


20.0 


20.0 


15.0 


15.0 


15.0 


5.0 


10.0 


10.0 


10.0 


25.0 


10.0 


5.0 


10.0 


10.0 


5.0 


10.0 


10.0 


5.0 


5.0 


10.0 


0.0 


5.0 


10.0 


0.0 


0.0 


10.0 


0.0 


0.0 


5.0 


0.0 


5.0 


5.0 


0.0 


5.0 


5.0 


5.0 


0.0 


0.0 


5.0 


10.0 


0.0 


5.0 


5.0 


0.0 


0.0 


5.0 


0.0 


0.0 


5.0 


425.0 


275.0 


440.0 



Rufous-sided Towhee (Pipilo erythrophthalmus) 
Solitary Vireo ( Vireo solitarius) 
Rufous-crowned Sparrow {Aimophila ruficeps) 
Canyon Wren (Catherpes mexicanus) 
Bewick's Wren ( Thryomanes bewickii) 
Blue-gray Gnatcatcher {Polioptila caerulea) 
Black-chinned Hummingbird {Archilochus alexandri) 
Western Wood Pewee (Contopus sordidulus) 
Black-chinned Sparrow {Spizella atrogularis) 
Ash-throated Flycatcher {Myiarchus cinerascens) 
Western Tanager (Piranga ludoviciana) 
Black-headed Grosbeak (Pheucticus melanocephalus) 
Brown-headed Cowbird ( Molothrus ater) 
Broad-tailed Hummingbird {Selasphorus platycercus) 
Bushtit (Psaltriparus minimus) 
Scott's Oriole (Icterus parisorum) 
Ladder-backed Woodpecker (Dendrocopos scalaris) 
Lesser Goldfinch (Spinus psaltria) 
Warbling Vireo ( Vireo gilvus) 
Violet-green Swallow ( Tachycineta thalassina) 
Hermit Thrush {Catharus guttatus) 
Western Flycatcher (Empidonax difficilis) 
Elf Owl (Micrathene whitneyi) 
Cassin's Kingbird ( Tyrannus vociferans) 
Rivoli's Hummingbird [Eugenes ful gens) 
Virginia's Warbler ( Vermivora virginiae) 
Hepatic Tanager {Piranga flava) 
Total 



Avifaunal Aspects. — Twenty-seven breeding species were recorded in North 
McKittrick Canyon over the 3-year study period (Table 7). Seventeen (63%) 
of the 27 species were present as breeding birds each summer of the study. 
Ten species (37%) nested on the study plot only 1 or 2 of the 3 years. Twenty- 
three, 20, and 24 species were represented for 1972, 1973, and 1974, 
respectively. Density values ranged from a low of 275.0 pairs per 100 acres in 
1973 to a high of 440.0 pairs per 100 acres in 1974. In 1972 there were 425.0 
pairs per 100 acres. 

Biomass figures differed only slightly for the seasons 1972 and 1974 (Table 
8). Corresponding to the marked decrease in density of individuals in 1973, 
the standing crop biomass was 34% less in 1973 than it was in 1972 and 1974. 
Consuming biomass showed a decrease of 35% in 1973 compared to 1972 
and 1974. 



BREEDING AVIFAUNAS 



195 



8 * 

00 co 

M OS 



2 £ 

5 2 






U 



PQ S 
< 






© S 

O M 



U O g 





PQ 
U 


Q 
O 


u 


CO 








r~~ 








OS 








"■* 










QQ 


o 


<ji 




( ) 


o 


u. 




oa 


-" 


TO 



u s 



CQO o 
U O g 



5 -el -. 



r- oo r- cn r- 

m i-t os © © 

00 so as SO Tj" 

T+ Tt ^ CO 



© o © © © © 

in (^ oo h ri io 

<—i r~ «y> co as r— 

oo co r» as as 

— 1 —I Tt 



•n-o\oooosO(N*— ir- 

<o<No-HOsoor^'sO 

n in tj- o -h so as co 
<-o io co cn cn 



ooooooooo o 

^^•'dvo^d^fnod oo 

Oi « 0O ^) (N ^hvoC^- O 

<n m .-< r- -^ h m oo 



so 


OS 


00 


00 


o 


»H 


«*■ 


r- 


SO 


m 


cn 


o — * 


O 


o 


m 


00 
cc 

CO 


IO 

a-. 

CO 


OS 


o 

(N 


o 


o 

OS 

o 


CN 

CO 


CN 


fN 


O 

o 


a. 


© o 

CO 


O 


o 


in 

o 



fN 


<n as <n 


O 


in 


lO 


oo r- co 




1- 


-f 


-H CO U-) 




m 



ooooooooo 

vd^-«oco"coo'r-'oo 

as o oo oo oo in 
fN r- oo co *— i 



^1 


a 


!/■) 


DC 


in 


a 


O 

r- 


lO 

a 

CO 


Os 
OS 
CN 


>0 

o 

(N 


00 


©s 



X 


c 


CC 


o 


>o 


as 


CN 


c 


o 


CO 
CO 

t 


O 

CO 


O 

CO 


o 

fN 


00 

CO 


o 
o 


O 


o 


o 



o 



o 


O 


o 


o 


o 


O 


o 


o 


o 


o 


o 


o 


G 


o 


o 


o 


"t 


ri 


r- 


fN 


-r 


as 


00 


ri 


o 


SO 


sO 


cc 


*i" 


o 


o 


Tf 


o 


CO 


co 


CO 


'.TO 


CO 


CN 


CO 


00 


SO 


SO 


VI 


. — i 






<tf 


On 
fN 


£1 


^ 


m 


^ H 


as 
»n 


o 


CN 


— 


r- 


co 


fN 


' ' 






so 
in 



co r- as co fN 



CNVOWICO-— lOsC-COOO 



o -a > 

S2 2 

Oh > C 

^ o & _ 

■g u w -5 

cotj .S 

> <U D t_ 

V x 2 5 2 

3 C V ^ -§ 

o £ * 

o 



3 3 2iS S 

& & pq oq j 



C/3 



•g 








-3 










o 

3 


O 


o 
3 

r- 

£ 
u 

CO 


o 


O 

r 


O 
.3 
> 

| 

^T3 




03 

ao 

r3 


« 

2 
S 


1) 

a> 

M 

o 


.3 
> 

"o 


o 

C 

o 


bO 

u 

3 


3 


3 
o 

73 
3, 
<0 


^ c 

1 1 



196 



NEWMAN 



3! 



U © ft 



CO S 

8 2 



CQ O 
U © 
CO ^ 



BQ 2 <" 

S 2§ 



oa o 
u © 

CO "l 



2 > 



r- on o o co 



in o h w 



O 00 ^H 


co 


oo co 


to 00* 


rt- 


to 


W1 


Tfr CO 


cs 


r- o> 


ctn r-- 


<* 


o> 


o> 


CS 


as 


CS <N 




r~ 







© © o o © © 

© © ©' OS vd to 

On O rH to VO 

tO »0 «— i CT\ i— i 

^H CO 



© © © o © 

oo m' h oo o\ 

r- r- to >o to 

*£ VO m M 0> 



">* 


On 


o 


o 


^D 


o> 


o 


rj- 


i-H 


o 


to 


to 

CO 


CO 


o 


o 


»o 


lO 


CS 
CO 
CN 




<o 


o 





ooooo q 

cn © © ©' oo' o 

c~- to r- © 

CS CS ■* © 



© © © © 



to 


^f 


«-H C 


© 


^o 


00 


to 


© 


to 


CO 


CO 


CO 















„_, 












00 


















o 












c 






■«t 


'St 


*"1 


© 


© 


°) 


S/5 


»o 


'-J 


—^ 


CO 


q 


2 


cs 


cs 


c*- 


t^ 


© 


00 


© 


cs 


C 


00 


i-H 


to 


00* 


CO 


'EJ 


© 


d 


to 


<* 


1 — 1 


-* 




VO 


•"" 1 


r» 


to 


CA 


r~ 


© 


Q 


as 


o> 


CO 


ri 


1— 1 






r- 


T3 


CS 


cs 






c^ 


i— i 


|— 1 














C 












<r> 


















3 












£> 


















O 












B 


















6 












H 






© 


q 


© 


© 


o 


© 




© 


q 


© 


© 


© 




© 


q 


CS 


to 


•sd 


O"^ 


© 


cs 




'X! 


^d 


r-H 


00 


CO 




cs" 


cs 


r- 


cs 


c^ 


■— 1 




i-H 




r~~ 


r- 


to 


lO 


vo 




© 


o 


cs 


vo 


cs 


'-' 




CO 




^D 


to 


CO 


CS 


oo 




r-~ 


c-> 



oo io oo a oo 



CO V£> i— I OO 



u 


(0 


£ 








<l> 






o 


£ 


O 


o 




_>> 


a) 


rt 




T3 


E 


Ph 


£ 


u 


.a 


Td 


m 


03 


X> 


T3 







o 


M 


<U 


o 


c 


>. 


g 





£ 


0) 


fe 


3 


c 


Sj 


c 


trt 


■£ 


<u 


-5 


OJ 


"g 



r^T .a C3 



3 £ £ :S S 
< ^ > ^ U 



O o ^ '2 



■a 

c_ 
-a 
o 
o 



6 r£> 



BREEDING AVIFAUNAS 



197 



s§ 



CQ O 

U o 

{/2 l-H 



lO M 10 


cs 


OO 


OO t-h CO 


CO 


r- 


^H 


CN 


co 



cn o co 


lO 


CO 


co co r- 


CO 


Tf 


«-H »-h 


CO 







CQ 

u 


© 

o 


09 


rn 








r^ 








o-, 








*"" 










CQ 


o 


0) 




u 


o 


U 




J-, 


*" 





X 

u 


o 
o 


r- 








CQ 
U 


o 
o 



UO lO lO 



q 


q 


q 


q 


q 


<N 


rs 


m" 


r-' 


_4 


CO 


r- 


r^ 


r- 


CO 


*" H 






CN 


0\ 

o 






< 



q q 
to o 



q q q q 

>o <n o r^ 

vo r^ co 



co \r> co 
co co r-~ 



3 "S 

£°3 

I -s -a 

£ £ x> 

= £| 

•o a: £ 
g 73 £ 

i T! 3 

I 2 > 

S £ £ 



198 



NEWMAN 



I 



1 



||i --■ll 



A — Foliage Insect 
B — Aerial Insect 
C — Ground Seed 
D — Ground Insect 
E — Foliage Seed 
F — Timber Drilling 
G— Foliage Nectar 
H — Timber Searching 
I — Ground Predator 



FIG. 5. — Percentage contribution of foraging classes to standing crop biomass 
(SCB) and consuming biomass (CB) for North McKittrick Canyon. Bars represent 
1972, 1973, and 1974, respectively. 







4* 



mw: 



FIG. 6.— South McKittrick Canyon, Guadalupe Mountains National Park, Texas. 



BREEDING AVIFAUNAS 199 

Six foraging categories were represented in North McKittrick (Fig. 5). 
The foliage insect and ground seed foragers were of major importance each 
year in this area. 

Species diversity was 3.00 in 1972, 2.83 in 1973, and 2.95 in 1974. 

South McKittrick Canyon 

Hahitat Description. — The study plot in South McKittrick begins a couple 
of hundred yards south of the juncture with Main McKittrick and runs 
linearly about 1 .0 mi. ( 1 .6 km) to the site of the old Grisham-Hunter Lodge. 



TABLE 9. Density of breeding-bird populations in South McKittrick Canyon for 1972, 1973, 
and 1974. 



Species 

Solitary Vireo ( Vireo solitarius) 
Black-chinned Hummingbird (Archilochus alexandri) 
Bewick's Wren ( Thryomanes bewickii) 
Broad-tailed Hummingbird (Selasphorus platycercus) 
Western Wood Pewee ( Contopus sordidulus) 
Western Tanager (Piranga ludoviciana) 
Black-headed Grosbeak (Pheucticus melanocephalus) 
Rufous-crowned Sparrow (Aimophila ruficeps) 
Ash-throated Flycatcher (Myiarchus cinerascens) 
Canyon Wren {Catherpes mexicanus) 
Warbling Vireo ( Vireo gilvus) 
Black-chinned Sparrow (Spizella atrogularis) 
Violet-green Swallow ( Tachycineta thalassina) 
Lesser Goldfinch (Spinus psaltria) 
Blue-gray Gnatcatcher (Polioptila caerulea) 
Rufous-sided Towhee (Pipilo erythrophthalmus) 
Western Flycatcher (Empidonax difficilis) 
Brown-headed Cowbird (Molothrus ater) 
White-breasted Nuthatch (Sitta carolinensis) 
Scott's Oriole (Icterus parisorum) 
Bushtit (Psaltriparus minimus) 
Hepatic Tanager (Piranga flava) 
Mountain Chickadee (Parus gambeli) 
Hairy Woodpecker (Dendrocopos villosus) 
Virginia's Warbler ( Vermivora virginiae) 
Grace's Warbler (Dendroica graciae) 
Rivoli's Hummingbird (Eugenes fulgens) 
Ladder-backed Woodpecker (Dendrocopos scalaris) 
Elf Owl (Micrathene whitneyi) 

Blue-throated Hummingbird (Lampornis clemenciae) 
Cassin's Kingbird ( Tyrannus vociferans) 
Hermit Thrush ( Catharus guttatus) 
Total 



Pairs 


per 100 


acres 


1972 


1973 


1974 


33.3 


25.0 


25.0 


25.0 


37.5 


20.8 


25.0 


12.5 


20.8 


20.8 


8.3 


25.0 


16.7 


29.2 


29.2 


16.7 


20.8 


16.7 


16.7 


20.8 


16.7 


16.7 


16.7 


16.7 


16.7 


12.5 


12.5 


16.7 


0.0 


4.2 


12.5 


20.8 


25.0 


12.5 


16.7 


8.3 


12.5 


8.3 


12.5 


12.5 


4.2 


16.7 


8.3 


16.7 


16.7 


8.3 


12.5 


16.7 


8.3 


8.3 


12.5 


8.3 


4.2 


12.5 


8.3 


0.0 


4.2 


4.2 


8.3 


16.7 


4.2 


8.3 


12.5 


4.2 


8.3 


12.5 


4.2 


8.3 


8.3 


4.2 


8.3 


8.3 


4.2 


4.2 


12.5 


4.2 


4.2 


12.5 


4.2 


0.0 


4.2 


4.2 


0.0 


0.0 


0.0 


4.2 


0.0 


0.0 


4.2 


0.0 


0.0 


4.2 


0.0 


0.0 


4.2 


0.0 


333.6 


341.7 


400.2 



200 



NEWMAN 



3 
(X 

o 
ex 

a 



c 









u 




-a 


-i- 
r- 




on 


X, 


T3 


U 


aj 


t/j 


„ 




m 




r^ 


C/3 


ON 


03 




fc 


„ 


o 


r-i 






,o 


On 


ft 


1—1 


o 






O 


o 




ci 


C 


e 






>, 


-d 


r 


c 


-i 


rt 


U 


on 


^ 



co S 

< 



u2 o 



CO o 



co 2 

5 2 



CO o 
U o 

00. — < 



s§ 



CO o 
U o 
in * 



3^o 



u 

o 



ri ^ ti- m' io ri ^ t^ vd h io 
vooooi/Ncoioor-ONONC-- 
co en <^i cn co co -h 



ONOcoocNi/or^Tj-oov) io 
CN O m io on r-° co oc I o iiO cn on 
■^■oooon-r^oooo^HaN^- r» 



ON 


"<* 


SO 


wo -3- 


«-h 


o 


ON 


o 


on 


Tf WO 


^o 


nO 


CO 


CN 


T 


rN oo 


CO 


,—1 


m 


CM 


CN 


co 


— H 


O 


cn 










i — 



o 


lo 


■* 


oo 


CO 


O 


00 

wo 

^o 


-r 


cn 


o 

n 
CN 


o 

CO 


CO 

CO 



CNCNtJ-ONOOCOONCOCOOOO -H 



_; 


r- 


-r 


o- 


i ~o 


_; 


"fr 


r- cn o o 


^sO 


IO 


o 


1/0 


C 


so 


SO 


o 


r- co co wo 


^t- 


-t 


co 


CO 


CN 


<-H 


»-H 


i— 1 




ON 



i— lOCNvooO^Or-'Tr^Owo^O 

r^ o r^ ^ wo" cn co oo o wo t}-' 

0N00CNC0COOO00r~-VO0N 

r^oocNwo^o v >ocN'-i 



*3" ON ON 0O CN CN 

o co" cn <—' xr co 

cn oo tj- r- co wo 

CN CN ^-i r- 



O TT WO CN CO 



oc. 


oc 


r- 


-t 


r-^ 


to 


wo 


■ — 


O 


*-t 


f- 


O 


so 


CO 


o> 


-t 




CN 



CN CN tO CN rt ^O 



.— i CO 00 CO 



C CNON^VOtJ-i— iCNONCNOwo 
►— ' VOOOCCNOOOOWOCOCOCOCN 



CO t^- CN .— I 



ONCNCOwor^ONCOCO^OWOON 
CN i CN UT) CN i — I Tt i-H wo' © wo r* 

^■c-oocncnooonc^^o^ 
Tt ^h on co co co ^h 



*o wo on co co *— i^trj-oo 
r»" on cn oo vd vd »-h oo' r^ 

<— I CN »— I CO CO r^ 



rf r~ co \o on 

o in -h oo' — < 

CN ^sO VO CN O 



O h VD io O CN 
00 ON CN CN O CN* 
IO CN O CO CO IT) 
\0 \Q KD CO CN -^f- 
CN 



r- ON CO CO CN 



-'-J 
c 


o 


60 


u 


CD 

73 


rt .a 










— I 




O 


«J 


O U 


cd 


^ 




73 


o 

3 


^ 




o 
o 


H 


>» s 


^ 


to 




c 


- 

CZ3 




o 

a. 


2 a 

.2 ° 
co S 


CO 

O 

- 


cd 

'Si 

> 


C/3 


- 

OO 



o 


T3 




^ 






73 


^) 




o 






p. 


O 

U 

-3 


X 


M 






00 

-o 


c 
H 


—5 






^ 


<u 


T-l 


(U 






-n 


C1J 


T 


TJ 


-3 

o 


u 


73 

0) 


U9 


g 


C 

O 
<u 


o 

s 


O 


3 
o 

3 


o 

-3 


3 

OO 


^ 


PQ 


pi 


C2 


hJ 





BREEDING AVIFAUNAS 



201 



s!§ 



MO S 

U o g 
oo ~ a 



si 



POO S 



s§ 



M O 

u © 

co *-< 



3 *&b o 



0) 

co 



co o>\ vo o-\ o 

m t»V oJ d 

cn oo co »-h 

N (N H H 



r- 


r-^ q q q 


«-H 


in q 




Tf O^ d d 


CO 


CN o 
CN 



ooo>o 

in o o r-~ 
o\ co i— o 
r- t— co cn 



On On 

co 



q q 
o o 



o o 

o o 



CO 


CO 


~h CT\ C-^ 


o 


C) 


oo 


On r- os 


CO 


CN 


ri 




r^ 



VO 


o o^ 


f» 


^1 


o 


O CTn 




O 



q 


q 


t> 


"1 


"1 


r- 


q 


o 


CO 


r- 


CN 


°°. *-? 


CO' 


to 


d 


«o 


l> 


,-H 


<j\ 


d 


d 


Tt 


ms 


,-H 


^h d 


_J 


On 


co 


<* 


o> 


o 


\D 


«t 




o> 


.— i 


lO 


o 


o 


r- 


r- 


CN 


'" H 


"<t 


co 

CN 


fN 




CN 


CN 


r- 


o 


O 



^r 


CI 


VO 


vo 


q 


CO 




CO 


O 


O 


q 


CO 


13 


q 


G) 


>^j 


CO 


I/-) 


t> 


©> 


d 


d 


C 


o> 


>o 


d 


d 


Tf 


'J5 


CN 


Os 


<~i 


o> 


^ 


CO 


c*» 




oo 


1—1 


3 


ir-j 






v£> 


Q 


i — i 


t> 


o> 


CI 


i — i 


1 — I 






v£> 


-c 


CN 


— 






co 


i — i 




r— 1 














c 

3 












^5 




















o 












£ 




















O 












H 








— J 


«o 


q 


tr, 


q 


^H 




q 


■tf 


© 


q 


TT 




9\ 


OO 


r- 


cn 


r^ 


d 


r-- 


d 


r» 




d 


l> 


d 


d 


t> 




d 


"<* 


^ 


MD 


i — i 


C- 


C^ 




*t 




CO 


t- 






W) 




o 


<y\ 


O 


o 


rt- 


co 


— 




o 




-3- 


ro 






OO 




IT) 


CN 


CO 


— 1 










CN 























00 lO 00 CT\ 00 


q co rH co 


-* rH 


h ri <t h t^ 

CO ^H -H — 1 Tf 


C7\ <— i u-> »o 

-h CO CN 


d to 

V.O co 



O <D > 

cfl > O 



<L> 



< £ 



d ^5 T3 
CO «J 43 



3 s 
> ^ 



00 


^ 




H 


% 


c 

o 


1 




■£ 




o 

Cm 


1 


(U 


r3 


u 


CQ 


u 


w 


EC 



o 

Ph 

o 

g T3 

cv <D 

H 2 



CO 



202 



NEWMAN 



S§ 



CO © a> 





CO 

u 


o 
o 


50 

o 


m 








r-- 








0> 








•^ 










C£ 


o 


V3 




u 


o 


u 




c/5 


• H 



g §3 



VO io VO O 

co n' ai d 



(N <N — 



CO 


o 


m o 


vo 


<* 


^_ 


—< 


CO 


c i 


«-H o 


00 

r- 

co 


© 


d 

wo 


CO 

"fi- 
ve 



r- co o ^o 


vq 


q 


q 


O) 


o> r-' o" d 

lO CO CO 




d 


d 


ON 



in oo 



q t> 

ON d Tj-' 











o 


LTy 


VO 


■o 


q 


r~; 




— i 


— 1 


q 




CQ 


o 
o 


CD 


0) 


VO 


CO 


cr' 


d 


o> 


S3 


CO 


CO 


MD 




U 


8 


z 


o 


OS 


rs 




CN 


o 


o 


r- 






cs 


CD 


1 — 1 








<N 


CO 


1 — 1 


i-H 


00 










M 












,_, 






CO 










.2 












CD 








cs 




















,£> 




























s 








•"■< 




















H 


















q 


CO 


CO 


o 


rH 




i-H 


^H 


r- 




BQ 


o 


a> 




irj 


o-" 


^H 


d 


^6 




C-« 


r^- 


_1 




r o 


© 


o 




VO 


"*■ 


VD 




tr- 




OS 


CT\ 


o 




on 


<"■< 




-- 1 


«— 1 






co 




cs 


CS 


r- 



CO ^o CO r— 

co co c-^ t> 



-a 


T3 




X) 

.a 




S 


IS 




X) 

s 
s 




a 


E 


X> 


3 




3 


e 






3 




ffi 




T3 


E 


£ 


T3 




CD 

C 

s 






CD 
cd 


rd 


2 


'ed 


O 


o 


o 


V 


B 5 * 


Js 


Jq 


Ji 


-a 


s 


*f 




o 


etf 


o 


o> 


00 


cd 





> 


3 







"rt 


03 


O 


O 

H 


-9 


.S 




<d 


3 




."£ 


CO 





CQ 03 rt pQ 



BREEDING AVIFAUNAS 



203 



The stream in this canyon flows above ground in segments and provides a 
fairly constant water source for wildlife. Elevation at the north end is 5200 ft 
( 1 585 m), increasing to 5300 ft ( 1 6 1 5 m) at the south end of the plot. As in the 
other two McKittrick plots, steep canyon walls rise abruptly 2000 ft (610 m) 
above the canyon floor. 

The vegetation in South McKittrick Canyon is dominated by chinquapin 
oak and big-toothed maple (Acer grandidentatum); importance values were 
79.25 and 75.50, respectively (Tables 1-4). Ten species were represented 
among 180 trees sampled. Common shrubs and grasses contributing to the 
understory in South McKittrick were Acer grandidentatum, Juniperus 
deppeana, Quercus undulata, Agave neomexicana, Cladium jamaicensis, 
Bouteloua curtipendula, and Muhlenbergia sp. 

Figure 6 is a view of this study plot. 

Avifaunal Aspects. — The breeding avifaunas in South McKittrick Canyon 
were represented by 32 species over the 3-year study period (Table 9). 
Twenty-four (75%) of these species were present each summer. The total 
number of species was 28 in 1972, 28 in 1973, and 27 in 1974. Density values 
ranged from a low of 333.6 pairs per 100 acres in 1972 to 400.2 pairs per 100 
acres in 1974. In the summer of 1973, 341 .7 pairs per 100 acres were recorded. 



C.B. 
% 



\V «H"h 



A — Foliage Insect 
B — Aerial Insect 
C — Ground Seed 
D — Ground Insect 
E — Foliage Seed 
F— Timber Drilling 
G — Foliage Nectar 
H — Timber Searching 
I — Ground Predator 



FIG. 7. — Percentage contribution of foraging classes to standing crop biomass 
(SCB) and consuming biomass (CB) for South McKittrick Canyon. Bars represent 
1972, 1973, and 1974, respectively. 



204 NEWMAN 

Standing crop biomass increased by 12% in 1973 compared to 1972, and 
by 11% in 1974 compared to 1973 (Table 10). Consuming biomass increased 
by 8% in 1973 compared to 1972, and by 16% in 1974 compared to 1973. 

Figure 7 shows the percentage contribution of each foraging category to 
total standing crop and consuming biomass for South McKittrick. The 
foliage insect feeders contribute by far the greatest biomass of the seven cate- 
gories represented. 

Species diversity values were 3. 15 in 1972, 3. 1 1 in 1973, and 3.20 in 1974. 

Upper Dog Canyon 

Habitat Description. — Upper Dog Canyon lies about 3.0 mi. (4.8 km) due 
west of the McKittrick Canyon system and is separated from that complex 
by mountainous terrain reaching 7700 ft (2347 m) in elevation. Drainage in 
Upper Dog Canyon is in a northerly direction. This canyon is dissected by 
the Texas-New Mexico state boundary. The study plot here runs from about 
600 ft (183 m) south of the Dog Canyon Ranger Station along the canyon 
floor a linear distance of about 1 .0 mi. ( 1.6 km). At the south end of the study 
area is a spring seepage, the only constant supply of above ground water on 
this plot. Elevation ranges from 6300 ft (1920 m) to 6700 ft (2042 m). 

Tables 1-4 summarize quantitative data respecting trees in Upper Dog 
Canyon. Two hundred trees belonging to 10 species were sampled. 
Chinquapin oak dominated the canyon woodland (importance value, IV 
108.06) followed in importance by big-toothed maple (IV 76.46). Common 
shrubs and grasses contributing to the understory in Upper Dog were Acer 
grandidentatum, Juniperus deppeana, Agave neomexicana, Opuntia im- 
bricata, Muhlenbergia sp., and Stipa tenuissima. 

The physiognomy of this study plot is shown in Fig. 8. 



' fa 





FIG. 8.- -Upper Dog Canyon, Guadalupe Mountains National Park, Texas. 



BREEDING AVIFAUNAS 205 



TABLE 1 1. Density of breeding bird populations in Upper Dog Canyon for 1972, 1973, and 
1974. 



Species 



Pairs per 100 acres 



1972 


1973 


1974 


37.5 


33.3 


41.7 


25.0 


29.2 


29.2 


25.0 


16.7 


167 


20.8 


16.7 


25.0 


20.8 


16.7 


25.0 


20.8 


16.7 


20.8 


16.7 


20.8 


12.5 


16.7 


16.7 


16.7 


16.7 


16.7 


16.7 


16.7 


16.7 


12.5 


16.7 


12.5 


8.3 


16.7 


16.7 


4.2 


12.5 


16.7 


16.7 


12.5 


12.5 


16.7 


12.5 


12.5 


12.5 


12.5 


8.3 


8.3 


12.5 


12.5 


4.2 


12.5 


8.3 


0.0 


12.5 


0.0 


0.0 


8.3 


12.5 


12.5 


8.3 


12.5 


8.3 


8.3 


12.5 


8.3 


8.3 


4.2 


4.2 


8.3 


8.3 


8.3 


8.3 


16.7 


20.8 


8.3 


4.2 


4.2 


4.2 


8.3 


12.5 


4.2 


16.7 


8.3 


4.2 


4.2 


8.3 


4.2 


4.2 


4.2 


4.2 


4.2 


4.2 


4.2 


0.0 


8.3 


4.2 


0.0 


4.2 


4.2 


0.0 


0.0 


4.2 


0.0 


4.2 


1.0 


1.0 


0.0 


1.0 


1.0 


1.0 


0.0 


12.5 


20.8 


0.0 


0.0 


8.3 


0.0 


0.0 


8.3 


0.0 


0.0 


8.3 


0.0 


0.0 


4.2 


435.5 


423.2 


459.4 



Western Wood Pewee (Contopus sordidulus) 
Solitary Vireo ( Vireo solitarius) 
Black-chinned Hummingbird {Archilochus alexandri) 
Chipping Sparrow (Spizella passerina) 
Warbling Vireo ( Vireo gilvus) 
Lesser Goldfinch (Spinus psaltria) 
Blue-gray Gnatcatcher (Polioptila caerulea) 
Western Tanager (Piranga ludoviciana) 
Ash-throated Flycatcher (Myiarchus cinerascens) 
White-breasted Nuthatch (Sitta carolinensis) 
Mountain Chickadee {Parus gambeli) 
House Finch (Carpodacus mexicanus) 
Virginia's Warbler ( Vermivora virginiae) 
Grace's Warbler {Dendroica graciae) 
Black-headed Grosbeak (Pheucticus melanocephalus) 
Western Bluebird (Sialia mexicana) 
Rufous-sided Towhee (Pipilo erythrophthalmus) 
Violet-green Swallow ( Tachycineta thalassina) 
Pygmy Nuthatch (Sitta pygmaea) 
Canyon Wren (Catherpes mexicanus) 
Brown-headed Cowbird (Molothrus ater) 
Rufous-crowned Sparrow {Aimophila ruficeps) 
Rock Wren (Salpinctes obsoletus) 
Black-chinned Sparrow (Spizella atrogularis) 
Western Flycatcher (Empidonax difficilis) 
Acorn Woodpecker (Melanerpes formicivorus) 
Bewick's Wren ( Thryomanes bewickii) 
Broad-tailed Hummingbird (Selasphorus platycercus) 
Hairy Woodpecker (Dendrocopos villosus) 
Hepatic Tanager (Piranga flava) 
Common Flicker (Colaptes auratus) 
Hermit Thrush (Catharus guttatus) 
Scrub Jay (Aphelocoma coerulescens) 
Steller's Jay (Cyanocitta stelleri) 
Cassin's Kingbird (Tyrannus vociferans) 
Cooper's Hawk (Accipiter cooperii) 
Great Horned Owl (Bubo virginianus) 
Bushtit (Psaltriparus minimus) 
Scott's Oriole (Icterus parisorum) 
Yellow-rumped Warbler (Dendroica coronata) 
Rivoli's Hummingbird (Eugenes fulgens) 
House Wren (Troglodytes aedori) 
Total 



206 



NEWMAN 



O 



CQ 




u 








c/1 




C/5 




73 




s 













,o 




M 




C 








P 
















on 




c 




o 




o 




-n 




c 




?3 








CO 




U 




i/i 




v -»' 


** 




r - - 


C/5 


9\ 


h 


•a 

c 


X) 


c« 


a, 


m 


o 


r- 




cr> 






OX) 




a 


(N 


-a 





r i 


c 


— i 


o 




>1 


UJ 


c 


-j 


03 


ca 


u 


< 




H 





u S 



ca o s 
U o g 



§ I 



PQ o S 



03 



MO S 
U o C 



2 > 



&o 



*-h lO OO CO I/O VO O 

h rt M N t ' 

p- oo m w> r 

(N (S M r) 



^ lO VO D CO OO CN 

6 co oo n ifl ri vo — h 

CN r- CN CN CO vo CO 
CN 



OCOCOOCNiC^O v *Ol/~>i/'>-— ■— 'CO 



JZ 



Oioc— locTN^-oocNOr-Tfi/') oi 

OOOOr^Tj-OOO OOuo^DcOr-HCO -H 

OONO^D'-^CO C) ^ (N (N C) M3 O 



^Hiocovor-^DOoir-^orJOO 



i-H Tfr CO CO MD — I 

r- oo in >o -< oo 

m <N m h ph 



Oc0000>O0"NOO00i/->ir)OO 

Oior^-OiorJ-OOcoOrNOO 

co co cn co co o t— i in ^ t 
O 0> O ^ CN co (S D (N h 



^Hio<No,a\<or~-<NONCoooo 
«-! ^ i> o> »o *h co \6 Tt t-h © © © 

>M O O V)0O(N ^ O* 
(N (N M M -h — i ^h 



Oc0O^0000NOOr~-OOOO co 

CJiOO^O^t-i^OcOioOOO <— i 

COOOOOcOooOiO'-hOCTn co 

oosooin^^rscJNH o 



.-1 


in 


VO CT\ '* CO CN 


-5fr rH CO r- CTn 


CO 


CO 

-+ 


r ) 


r- cn t-t *0 O 

,_| r- 1 r-H CO — H 


co ^o r- io cn 


cc 
CO 



C3 • 



H > 



r3 

£6 



u> ^ Jj <4 






oc .£ ^ 2 

H?! Its?* = 

ca •£ c/o •£ 



C/3 



2 .2 ~ 

bo C HM 

<u 'Eh 






EixOffl>ca>w 



BREEDING AVIFAUNAS 



207 



fflo S 
u 2 8 



MO S 

u o 8 





CO 

U 


O 
O 


W2 

o 


T3 

3 
3 


<n 








O 










O 


'"" 












CQ 


c 


C/3 






! ) 


© 








C/5 


"* 


CS 





0) 

ft 



^o >n r- m in ^f 



ioooa\tO\0^(N 



Tf o on cs r- o >o 
o r- cs oo cs cs co 

co rH *0 co co CS O 



o> 


V© 


in 


MD ON 


^O Tt 


CT-' 


00 


or. 


ro 


rr, 


Tf <N 


— < o 


ro 


ON 


<-h 


— 


CS 


-—i 


1— 1 I— 1 




On 



CS 


CO 


o 


CS 


co 


W> 


o 


OO 


X 


-* 

>/-) 




o 
oo 
in 


to 

CS 


co 
cs 


rs 

oo 
cs 


o 

(S 


in 


o 

ro 



CN Tf ON ^O Tf U"> 

Tf CS -t m vo oo 

(N (N M ^ h 



io ro in 00 



r- oo r- r-~ cs o co 
o r» n- o on rs io 

ON *«D ON CO Tj- (N vo 

ro 



lOOOvo-^-OOinO 



(NOOtMOi^'tO 

t tV to d m' o\ d 
in ^o oo cs OO lO 

^ -^- CO r-H CS ^ 



3? on vo r~- uo m Tf 

cs ^ i/-> on on m 

Tf- CS v£> VO O 00 

(N fS H H H 



in O — h r-- © oo 

r- oc i on cs i r-° d 

O h(N 00 M fS 

ON \D ^D co ro cs 



CO CO ON CS t— CO 



^D 


QJ 


«n 


— < 


CO 


00 


r-~ 


o 


cs 


o 


V£> 


r~ 


c 


o> 


VO 


v£> 


_ 


IT) 


p- 


iv) 


- 


r-| 


ON 


"~ ' 


CO 


o 


©- 


ON 


^O 


r~ 


ro 




^O 


ON 


T3 

C 
3 
O 




cs 














CO 



6 



cs 


o 


v£> 


CO 


r- 


<o 


sO 


O 


o 


"t 


o 


r i 


r- 


sG 


r- 


o 


o 


ON 


in 


o 


CO 


^ 


1 — 1 


cc 


CO 




o 


i — i 


r^ 


Tf 


cs 


cs 


r— 1 






o 


i — 
















m 



rf-OMDONOOCOvO'^- 











Z 








•a 




O 


!» 
















,o 




S 


o 


-C 




o 
U 

T3 




ft 






H 




o 

c 


CO 

-a 

O 

C 


GO 

■n 


0) 


43 


0) 


Tl 





3 


3 




'J 




a 


g 
2 


3 




c 




a 




O 


3 


o 







■a 


- 


O 




3 


^ 


cc: 


as 


CQ 


-J 


CC 


CQ 



r - 3 



CQ 



S •£ ^ - 

^ ,3 O <U 



3 ^ > 
O o 0> 

3 £ 3 

O 



uailuffl 



208 



NEWMAN 



U S y 



88 2 

35 - a 






CQ O 

u o 

W3 "-* 



Si 



PQ O a) 



§^ 



00 



-t- 


^o 


«-H O 


iy> 


v£> 


^n 


C 


"<fr 


o> 


co 




c-» o 

on 


ON 
ON 


o 
o 


<N 


o 


on 


o 
rs 



^H 


IT"; 


>o 


o 


o 


^H 


o 


O 


r- 


o 


-t- 

o 


o 


o 


to 


o 
o 


cn 


o 


CO 



-* 


i/1 


o 


-f 


CS 


VN 


to 


o 


o 


oo 

ON 

c) 


on 

ri 

m 


o 


o> 


O 


ON 

oo 


<N 


o 


o 



<— i io O c~~ W-) 


oo 


o o o 


o 


ON 


ON 


■<f 


CN CN o »o r> 

^> CO -rj- ON 

o oo in n 


CO 
WO 


cs o o 

liO 


cn 

IO 


r- 

on 
to 


r- 

On 

tn 


r- 

O 















T3 










«t 


q 


l — 1 


v£5 


MD 


l> 




ri 


00 


^ 


"* 


CO 


,— I 


r-' 


C-^ 


ON 


CO 


00 


O 


IO 


ON 


VT* 


ON 


r- 


ON 


CO 


r- 


00 


<u 


~f 


LO 


CN 


CN 


r) 


on 








ON 


00 

'o 
b 


<~! 






^ 


— 


"1 


Wi 


q 


«o 


V£> 




»o 


r- 


C> 


ON 


CN 


r-' 


rH 


© 


r^ 


oo' 




d 


\6 


r-H 


00* 


^D 


m 


o 


r-> 


ON 


VsO 




cn 


-* 


ro 


o 


o 


ON 


-t- 


co 


i — 1 


ON 




i — i 


00 


v£> 


MD 


— 










<N 




<— 1 






CN 



60 










C 






to 




£3 
o 


«* 


^t 


c 


NO 


R) 


t-^ 


c-^ 


•pj 


«N 


tu 


o 


o 


Q 


— H 


00 


(N 


CN 




1-1 








u 




£ 










H 






H 






ON 


ON 




-t 




t*^ 


r- 




t-^ 




ON 


ON 




o 



00 lO 0O 0O ON 



^h rN r- rr rH 


OO O WO 


fO rH '■t rH r- 


^O O f- 



a <d 

03 > 

a > 

T3 O 

<U O 

£ c 

^ 5 . 5 



£ 








1 








O <-i 




»H 












5 <u 




<u 




3 








ed 


Acorn Woodpeck 
Steller's Jay 
Scrub Jay 


"3 


D 
U 

Jo 


"c3 


o 

o 


"c3 




O 
2 


o 
2 


o 


o 
2 


3 


3 


<u 


3 


>, 


3 


V2 — 1 "S 


00 


OO 


."3 


OO 


.3 


lo 


8 !? 1 
u > ^ 










rt 

X 





BREEDING AVIFAUNAS 209 



8 2 

— o 



CO o 

U o 

05 «h 



V « 



CQ O 
U O 
C/3 -* 



CN 


00 


(N 


<N 


r- 


d 


o-' 
m 


<N 


OS 





-H O 


(N 


«o 


>o d 


vd 


CN 


r- 


-* 


CTs 




~ 





<N <N 


o 


^f 


ro 


d d 


d 


d 


en 


r-H r^ 




m 


uo 






<N 







s§ 


VI 


o 


>o as o 




0> 


CD 


vd <x> d 

O T-H 






OS 




*■"* 


rs 






•2 












O (N O 




PQ O 
U O 
00 — i 


a) 




in d c*> 



if § 

23 S w 



m vo c-> 

m ri c^ 



3 "2 

X 5 -S 

-a E e 



o 



-o - 

0) 3 



S ° 3 

2 ^ E o 

° 5 ° oo 

^ o > ^ J 



210 



NEWMAN 



C.B. 
% 




A — Foliage Insect 
B— Aerial Insei t 
C — Ground Seed 
i) — (.round Insect 
t — Foliage Seed 
F — Timber Drilling 
G — Foliage Nectar 
H — Timber Searching 
I — Ground Predator 



|l| H| "■ Hi 



FIG. 9.— Percentage contribution of foraging classes to standing crop biomass 
(SCB) and consuming biomass (CB) for Upper Dog Canyon. Bars represent 1972, 
1973, and 1974, respectively. 

Avifaunal Aspects. — A total of 42 species utilized the Upper Dog Canyon 
plot for breeding over the 3-year period (Table 11). Thirty of these species 
(71%) were represented each summer of the study. Total number of breeding 
species present each summer was 37 in 1972, 33 in 1973, and 38 in 1974. 
Density values fluctuated from a low of 423.2 pairs per 100 acres in 1973 to a 
high of 459.4 pairs per 100 acres in 1974. In 1972 there were 435.5 pairs per 
100 acres. 

Standing crop biomass and consuming biomass were lowest in 1973 
(Table 12). Standing crop biomass was 17% greater in 1972 than in 1973, and 
12% greater in 1974 than in 1973. Consuming biomass was 11% greater in 
1972 than in 1973, and 10% greater in 1974 than in 1973. 

Of the eight foraging classes represented in Upper Dog Canyon, the 
foliage insect category contributed the highest biomass (standing crop and 
consuming) percentage (Fig. 9). 

Species diversity values were 3.40 in 1972, 3.37 in 1973, 3.43 in 1974. 

The Bowl 

Habitat Description. — The Bowl contains a conifer-dominated community 
(Fig. 10) of about 300 acres (121.4 ha) located at the top of Pine Top 
Mountain. It lacks the weather-moderating influence exerted on the other 
study areas by steep canyon walls. Gentle slopes lead up to the rims of The 
Bowl from a low elevation of 7750 ft (2362 m) to 8000 ft (2438 m). Situated in 
The Bowl is a man-made impoundment created when the land was in pri- 



BREEDING AVIFAUNAS 211 

vate ownership. This tank is usually dry during the months of June and July, 
depending on precipitation. When it is dry, wildlife must obtain drinking 
water from one of the spring-fed streams of a nearby canyon. The study plot 
in The Bowl consisted of 24 acres (9.7 ha) situated away from major dis- 
turbance areas (e.g., log cabin, surface tank, meadow). This plot was 
rectangular in shape and ran in an approximately east-west direction along 
the gentle slope of the south side of The Bowl. 

Tables 1-4 summarize data descriptive of the forest on this plot. Limber 
pine (Pinus flexilis), ponderosa pine {Pinus ponder osa), and Douglas fir 
(Pseudotsuga menziesii) all show high importance values (98.22, 80.25, and 
67.65, respectively). Six species were represented among the 240 trees 
sampled. The understory in The Bowl consists primarily of young conifers 
{Pinus flexilis, Pinus ponderosa, Pseudotsuga menziesii) with some Quercus 
gambeli and a grass cover of Muhlenbergia sp. 

Avifaunal Aspects.— Total number of breeding species represented in The 
Bowl study plot over the 3-year period was 32 (Table 13). Of this total, 21 
(66%) were present each summer. Numbers of breeding species per year were 
25 in 1972, 26 in 1973, and 26 in 1974. Density values were relatively con- 
stant— 237.5 pairs per 100 acres in 1972, 244.8 pairs per 100 acres in 1973, 
and 236.3 pairs per 100 acres in 1974. 

Standing crop biomass and consuming biomass fluctuated slightly from 
year to year over the 3-year period (Table 14). The lowest standing crop bio- 





<&**- '" t '*%* "L ■ 




4$ 


-•• &«* * \i 




■< % ■ 


/**•/ "W^ M 




: *& 


k ^ ; iJfiS • „jL 


'- -* 




""* ~% . '"'-> % * ' °J&T* 


; ^"- '4 








; V* • 


. if* % *jl I *"♦ ■ 








Fig. 10. — The Bowl, Guadalupe Mountains National Park, Texas. 



212 NEWMAN 

mass was in 1973. In 1972 standing crop biomass was 9% greater than in 1973 
and in 1974 it was 15% greater than in 1973. There was a 4% greater con- 
suming biomass in 1972 than in 1973, and in 1974 it was 5% greater than in 
1973. 

Nine foraging categories were represented in The Bowl study plot (Fig. 
11). The foliage insect category contributed the highest biomass (standing 
crop and consuming) percentage. The aerial insect category was greatly 
reduced in importance in The Bowl as compared to its contribution in the 
other study areas. 



TABLE 13. Density of breeding bird populations in The Bowl for 1972, 1973, and 1974. 



Species 

Gray-headed J unco (Junco caniceps) 

Mountain Chickadee (Parus gambeli) 

Pygmy Nuthatch (Sitta pygmaea) 

Yeliow-rumped Warbler (Dendroica coronata) 

White-breasted Nuthatch (Sitta carolinensis) 

Solitary Vireo ( Vireo solitarius) 

Red Crossbill (Loxia curvirostra) 

Orange-crowned Warbler ( Vermivora celata) 

Rufous-sided Towhee ( Pipilo erythrophthalmus) 

Black-headed Grosbeak (Pheucticus melanocephalus) 

Western Flycatcher (Empidonax difficilis) 

Steller's Jay (Cyanocitta stelleri) 

Hermit Thrush (Catharus guttatus) 

Warbling Vireo ( Vireo gilvus) 

Hairy Woodpecker (Dendrocopos villosus) 

Common Flicker (Colaptes auratus) 

Western Tanager (Piranga ludoviciana) 

Grace's Warbler {Dendroica graciae) 

Broad-tailed Hummingbird {Selasphorus 

platycercus) 
Violet-green Swallow ( Tachycineta thalassina) 
Brown Creeper (Certhia familiaris) 
Western Bluebird (Sialia mexicanus) 
House Wren (Troglodytes aedori) 
Western Wood Pewee (Contopus sordidulus) 
Blue-throated Hummingbird (Lampornis 

clemenciae) 
Acorn Woodpecker (Melanerpes formicivorus) 
Whip-poor-will (Caprimulgus vociferus) 
Flammulated Owl (Otus flammeolus) 
Saw-whet Owl (Aegolius acadicus) 
Great Horned Owl (Bubo virginianus) 
Cooper's Hawk (Accipiter cooperii) 
Red-tailed Hawk (Buteo jamaicensis) 
Total 



Pairs 


per 100 


acres 


1972 


1973 


1974 


33.3 


25.0 


20.8 


25.0 


25.0 


20.8 


16.7 


16.7 


16.7 


12.5 


20.8 


20.8 


12.5 


8.3 


8.3 


12.5 


8.3 


8.3 


12.5 


0.0 


0.0 


8.3 


16.7 


12.5 


8.3 


16.7 


12.5 


8.3 


8.3 


16.7 


8.3 


16.7 


4.2 


8.3 


8.3 


8.3 


8.3 


8.3 


8.3 


8.3 


4.2 


12.5 


8.3 


4.2 


8.3 


8.3 


4.2 


8.3 


4.2 


8.3 


8.3 


4.2 


4.2 


12.5 


4.2 


8.3 


4.2 


4.2 


12.5 


4.2 


4.2 


4.2 


4.2 


4.2 


4.2 


4.2 


4.2 


4.2 


0.0 


4.2 


0.0 


0.0 


4.2 


0.0 


0.0 


0.0 


0.0 


4.2 


0.0 


4.2 


4.2 


0.0 


1.0 


1.0 


0.0 


1.0 


0.0 


0.0 


0.0 


1.0 


0.0 


1.0 


0.0 


0.0 


0.0 


1.0 


237.5 


244.8 


236.3 



BREEDING AVIFAUNAS 



213 



u s s 



pa © 

U © 

CO ^ 



81 



CQ O £ 

co -< £ 





aa 
U 


© 
o 


03 
y 

TO 


B 


cn 










r* 








O 


O", 








— 


*~* 












ca 


o 


BO 






u 


o 








CO 


*-" 


to 





|f3 



0) 

a, 

CO 



CN Tt <n in on <— i iO OO VO "O 

CO ^H i— I t— I CN *— I 



iA) On CN O rN 



o 


Tt ON 


On O CO 


*o 


Tt 


r- cn 

CO .— 1 


*>0 CN 


ON 



on 


ri 


r . 


l> 


'sD 


C-» 


m 


in 


CO 


O 


■o 


CN 

-i- 


Tt 

r- 
<* 


CN 

on 

CN 


O 
t 
co 


O 
CO 
in 


0> 
00 

Tt 


CN 
CN 
CO 




CN 


o 
cs 


o 
in 



00 


CO 


CN 


O 


O 


r^ 


o 


00 


in 


O 


VO 


O 


CO 


CN 


CO 




ON 


-t- 


CN 


-t 


n 






o 


CN 










CO 



©' co cn >/-> on 

00 CO © Tt O 

^H CN «-H ^H CN 



*— i©cNr-;v > or--Tt^o©vo 

r^ocNOvdoNoo'ooo 
>-<r~-ONTtcooo©ONr"-r-- 

C^inCNCOi/NTtr-iCNCO 



in ON CN CN CN © 
ON ON On On co ^^ 
00 CN VO CO CN IO 



CN CO CN 00 O to 

^t M vi vi \d ON 

IO CN CO ON ON o 

r-i Tt CN O 

— < CN 



OCOVOCOCN^OOOCOCNCO 

oocoinTj-cNc--oo^OTtco 



— j O © r-» in oc i «-h Tt 
r^ o O O cn r-' Tt Tt-' 

r-inTtcococNcN'— i 



93 


in 


On 


CN 


CN 


o 


b 


-r 


On 


-ON 


ON 


o 




t> 


CN 


MD 


CO 




-rs 


co 










B 












3 


























cn Tt ^o cn on m on 

CO i-h t~^ © CN ON CN 



r- oq 

00 Tt 



CO CO CN 00 O i— I 

©' 'oo in in O O 

00 CN CO ON Tt 

CN Tt CN O 

CN CO 



Tt 00 O Tf O 



to 






a) 

|1 

T3 



i-i _- to 

03 '~ 



§"3 

^1 si an 

w S ££><££ O > O 






J_ 




-a 




% 




B 


-5 




o 

T3 




3 


B 


a; 


— 
C 



£ 
Fi 


,B 

H 


02 
B 

W3 


03 

3 


TO 

jjj 
g 


n 


03 


cu 







U 


E 


p£ 


X 


u- 



214 



NEWMAN 









co 


O 


© 


CO 


o 


-r 


rf 




§ 


co 


d 


d 


CO 


© 


CN 


CN 


C3 


t> 






C- 


r- 


i — 


oc 


GO 


,_N 


^D 






VO 


o 


CN 


CN 








"■H 






«— < 


•— ' 


«"H 


r^j 



u a 









U 

0) 

n, 

GO 



O iO 


CO 


CN On 


T-H 


vo 


vq 


i— ( CO 


-t 


d *h 

CN 


ON 

cn 


On CN 
OO "*■ 
CN CN 


CN 

CO 

in 


CN 
(N 
CN 


CN 
CN 
CN 


CO CN 
© CO 


CO 



CO © 



© r» 


r- 


CN 


CN 


r- 


C 


o> © 


ON 


© 


© 


ON 


o- 


OO ON 


r~- 


© 


© 


CN 





© © 

© © 



r~ rj- cn 
tJ- cN r- 

CO CO VO 



l> © 

on r- 

CN 



T3 










T3 








2 




1) 


00 


00 


© 


«sO 




q 


CO 


CO 


i 


<q 


CO 


r^ 


d 


d 


od 


CO 


CO 


,— H 


^ 


S 


CN 


<u 


© 


co 




CO 


T3 


NO 


NO 


CN 


CN 


M 


CO 


CN 




l/-> 


C 


^f 


i-H 


NO 


,_, 


<^J 


rt 










3 








CD 




■_3 










o 








X> 




O 










M 








s 




fc 










o 




























H 






CO 


io 


© 


OO 




CN 


NO 


CO 




no 




CO 


r^ 


d 


© 




«0 


cn 


C- 




CN 




t-» 


CO 




, — 1 




CN 


© 


CN 




© 




NO 


CO 




»/N 




rf 


NO 


© 




© 




i-H 






CN 




i— i 




CN 




1 — 1 



Cfl in CN 

« «o co 



© CO 00 

O m ^ 



r^ d od 



On Tt 

r-^ oo" 





l_ 




o 


4= 










o 

c 

I—) 


O 






T3 




H 




a 




-o 


3 




,o 


O 


"e3 


<D 


Tj 




o 




*o 







£ 


o 


0) 


? 


C 


U 
-a 


c 

o 


3 
CO 


>> 


3 

«2 


2 
CO 


cu 


o 






3 




Pti 


< 




6 


ffj 








u 


a 


75 


u 


r3 


0) 


U 




5 


U 


T 




^E 


o 



^ PQ 



BREEDING AVIFAUNAS 



215 





SI 


o 
i- 
o 

P3 


m 

on 








ss 

C/3 ^ 


0> 

1- 

CJ 
C3 



s 3 

a> si? 



ex 



CO 


© 


•■H 


"* 


q 


ON 


On 


q 


q 


(N 


d 


d 


r-" 


I> 


d 


00 


00 


d 


d 


co 


-t- 




On 


CO 




" 


" 






CO 

i/-> 

CO 



O (N CN 



CN O ~h 

dor-' 



o m 
d r^ 



io q xr> 


q 


O oo 


00 


q 


q 


q 


r»' d — I 

On O 
co ■<*■ 


On 

On 


d on 


On 

I/O 


oo 


oo 


co 

MS 
NO 





CO 


O 
O 


0) 


CD 
C 


ON 


NO 


— 
d 


CN 


o 
Z 

(0 


NO 

d 


ON. 




it3 
<D 


o 
d 


q 
d 


*s- 




U 




r- 


*t 




CN 


en 


'" H 


<* 


£ 






On 
Ti- 










rt 










b0 








T3 






ro 




















c3 








C 








<^1 








<D 










;sj 








3 








r- 








< 










O 








O 








o> 
















tu 








IH 








— * 


























o 


















m 


q 


o 


«n 




r-; 


CN 


o^ 




q 


o 


ro 




CQ 


o 


a> 




r- 


u-5 


d 


cn 




<<* 


d 


«* 




d 


d 


OO 




U 


o 


u. 




On 


o 




O 




^o 


ro 


On 








OO 




00 


»-" 


C3 




— 


— 1 




CO 
















NO 
































CN 



r3 


T3 


3 




O 


O 



o 





c 


c 


o 








Cu 


5 


u 


<v 


&, 


2 


oo 


t« 


oo 



£ £ 



T3 

£•& 

£ i 

£ i 

3 £ 

-a K 
o -a — 

o ^ o 
£ 2 •£ 

V ^3 3 

| g « 
3 PQ 



rt 



73 



S (2 

3 
CO 



216 



NEWMAN 



Foliage Insect 
Aerial Insect 
C — Ground Seed 
D — Ground Insect 
E — Foliage Seed 
F--Timber Drilling 
G — Foliage Nectar 
H — Timber Searching 
I — Ground Predator 




Fig. 11. — Percentage contribution of foraging classes to standing crop biomass 
(SCB) and consuming biomass (CB) for The Bowl. Bars represent 1972, 1973, and 
1974, respectively. 



Species diversity values were fairly constant for each of the 3 years — 3.02 
in 1972, 2.98 in 1973, 3.03 in 1974. 



Avifaunal Comparisons 

Table 15 compares all breeding bird species recorded during 1972, 1973, 
and 1974, for each of the woodland habitats under consideration in this 
study. Of the 57 species represented, four (Solitary Vireo, Western Tanager, 
Black-headed Grosbeak, Rufous-sided Towhee) were successful in meeting 
their needs each year in all of the study areas. Six other species (Western 
Wood Pewee, Violet -green Swallow, Broad-tailed Hummingbird, Western 
Flycatcher, Hermit Thrush, Warbling Vireo) were recorded in all habitats 
under consideration, but did not occur each year in all areas. The Western 
Wood Pewee was a marginal breeding species in the conifer-dominated 
Bowl, being found there only in 1972. The preferred habitat of this species in 
the Guadalupes was that of the deciduous woodlands which provided more 
desirable nesting sites and more open canopy better suited to its feeding 
behavior. Violet -green Swallows were recorded nesting only 1 year in North 
McKittrick Canyon and 2 years in Upper Dog Canyon. A small number of 
suitable nesting sites and therfore increased competition by other hole- 
nesters for available sites may account for the fluctuations of this species. 
Broad-tailed Hummingbirds nested in all habitats studied but were present 
only in 1972 and 1974 in Main McKittrick Canyon. This species is probably 



BREEDING AVIFAUNAS 217 





r- 




W 










£ 




o 


m 


P3 


On 


0) 




P 


CS 




r- 




o 



& 



o ', 
c/5 - 



X X 

X X 

XXX 

xxxxxxxxxxxxxxxxxxxx 



X X 



xxxxxxxxxxxxxxx 









X X 


M 


»"H 






o 














XX XX 


s w 








« u 








s 


<N 










X 


X X 



xxxx xxxxxxxxxxxxxx 



xxxxxxxxxxxxxxxx 



x xxxx xxxxxxxxxxxx 



xxxxxxxxxxxxxxxx 



xxxxxxxxxxxxxxxxx 



XX XXX X xxxxxxxxxx 



XX XX XX xxxxxxxxxx 



x xxxxxxxxxxxxxxxxx 



xxxxxxxxxxxxxxxxxx 



XX xxxx xxxxxxxxxxxxx 



C3 
O 

X IX, a 
O > T3 .O 

1) O <D V5 

.a h 3 o 

s i-S § 

m o 5 



o 
o 



■as 

c o 



2 b W 



1 aS 

$ is 



M 03 

2 c 



o^ 






* I 



"5 



£ 
6 

" o 

- X g u & 

0> o> T3 _0> O 

c c o> T3 o 

.£ .5 3 g £ 

^ .* C | u 

f i f) -L "■* Ti 






> H 

c3 a) 






218 



NEWMAN 





** 




r- 




c^ 


M 


i— < 










rO 


^ 


t*» 


o> 


OS 


a 


l-H 


a. 




o 


<N 



o a 

CO » 







r- 






cr, 




u. 


>— < 




o 




1-1 


'B 


ro 

h- 


c 


2 


o> 


£ 


_ 




S 


r- 







-1- 






r~- 






OS 




.* 


»— I 




o 










g 


^ 


prj 


'3 




o> 


p: 








s 


On 



X X X X X X X X XXX xxxx xxxx 



XXXXXXXXX XXX XX xxxxx 



xxxxxxxxx xxxxxx xxxx 



XX xxxxxxxxxxx XXXX X 



xxxxx X x xxxx xxxxx 



xxxxxxx X xxxx xxxxxx X 



XXXXX X XX XXX 



XXXXXXX X XX X 



XXXXX X XX XXX 



XX XXX 



XX X 



xxxxxxx 



xxxxxxxx 



XXX 



xxxx 



o £ 

T3 -rt 



£ .5 

o P i- 

o — tu £ 



T3 Of, 



42 

GO 

c 



►H °° o 
Da '5- >, 



O Cu 



S-.b 



«j e =r — 



Si 

3 £ 

« c 

c o 

S3 £ 

tS £ 

« O 



£ 2 



o ^ 

§ES3 

si &S 



3 " "S 

V «> «j 

£ \* .e 

o « 4 



ffl^>m^i^3:s^uio^ffl^u<ou 



oo a 



NEWMAN 



219 



X X X X x 

X X X X X 









CN 


OD 


t-H 







Q 


rn 


53 


On 


a. 


i—i 


B, 




D 


<N 




r- 




CTs 








r^ 




on 


M 


«- * 


o 








rrs 


o> 


** 





X X 



x 



s * 

« o 

s 



TO 






_ 








"O 






£ 






^ 


0) 

C 

o 


(U 

ex 

CD 


3 
£ 


O 

-a 
S3 


IS 


"3 

o 


ed 






o 
o 


_£3 


O 

U 




TD 




U 

c 


3 
£ 
S 

ed 


2 


ed 


03 

6 


o 

1— 


s 




cd 


-a 



220 NEWMAN 

dependent on a food source provided by more mesically adapted flowering 
vegetation than is consistently found in Main McKittrick. The Western Fly- 
catcher, Hermit Thrush, and Warbling Vireo found the lower canyon 
systems suitable only as marginal habitats. These species were consistent 
breeders in the more mesic communities of the higher elevation. 

Four species (Blue Grosbeak, Olive-sided Flycatcher, Gray Vireo, Brown 
Towhee) were found only in Main McKittrick. Three of these species (Blue 
Grosbeak, Gray Vireo, Brown Towhee) were found more commonly near 
the mouth of Main McKittrick than on the census plot. As environmental 
factors fluctuate from year to year these more xerically adapted species may 
move, during the drier years, further into Main McKittrick. Olive-sided Fly- 
catchers were recorded consistently, during the breeding season in 1973 and 
1974, only in Main McKittrick. I list this species as a potential nesting species 
as I have no definite proof that it successfully nested. 

There were no species found only in North McKittrick or only in South 
McKittrick. One species (Chipping Sparrow) was unique to Upper Dog 
Canyon. There were several species which were found only in the coniferous 
forest community of The Bowl. These species were Gray-headed Junco, 
Orange-crowned Warbler, Brown Creeper, Whip-poor-will, Flammulated 
Owl, Red Crossbill, Saw-whet Owl, and Red-tailed Hawk. Eight species 
(Western Bluebird, Acorn Woodpecker, Common Flicker, Great Horned 
Owl, Cooper's Hawk, Pygmy Nuthatch, Yellow-rumped Warbler, and 
Steller's Jay) were found in both Upper Dog Canyon and The Bowl. The 
inconsistency shown in the data for owls may be due to difficulty in gather- 
ing owl data rather than a true indication of presence or absence. 

DISCUSSION 

The greatest avifaunal species diversity (total number of species and 
Shannon- Weiner index), density, standing crop biomass, and consuming 
biomass occurred consistently throughout the 3-year study period in Upper 
Dog Canyon (Table 16). The elevation and terrain of Upper Dog Canyon, 
being intermediate between that of The Bowl and the McKittrick Canyon 
system, allows for mixing of tree species creating a transitional zone between 
the two extremes. The greater structural diversity of the vegetation in Upper 
Dog thereby presents conditions allowing a greater avifaunal complexity. 

Stability of breeding avifaunas from year to year as judged by biomass 
computations, number of species, and species diversity was best maintained 
in The Bowl and Upper Dog Canyon communities. Due to terrain features, 
the vegetational aspects of these two communities are less likely to suffer 
flash flooding disturbances than are vegetational communities of the 
McKittrick system. The greater vegetational stability of The Bowl and 
Upper Dog Canyon communities accounts for the greater stability in 
breeding avifaunas of these two areas. Flash flooding in the McKittrick 
system creates dynamic vegetational communities along the canyon floors. 



BREEDING AVIFAUNAS 



221 



TABLE 16. A comparison of total numbers of species, Shannon-Weiner Index, density (pairs 
per 100 acres), standing crop biomass (SCB), and consuming biomass (CB), for breeding 
avifaunas of selected woodlands of the southern Guadalupe Mountains. 



Study area 


Total 


and year 


species 


Main McKittrick 




1972 


26 


1973 


25 


1974 


28 


North McKittrick 




1972 


23 


1973 


20 


1974 


24 


South McKittrick 




1972 


28 


1973 


28 


1974 


27 


Upper Dog 




1972 


37 


1973 


33 


1974 


38 


The Bowl 




1972 


25 


1973 


26 


1974 


26 



Shannon- 
Weiner 
index 



Pairs per 
100 acres 



SCB CB 

(g per (g per 

100 acres) 100 acres) 



3.05 
3.04 
3.16 

3.00 
2.83 
2.95 

3.15 
3.11 
3.20 



3.40 
3.37 
3.43 

3.02 
2.98 
3.03 



253.1 


9691.0 


3096.0 


206.6 


9365.6 


2819.4 


256.5 


9660.8 


3117.3 


425.0 


16697.0 


5322.7 


275.0 


10981.0 


3468.5 


440.0 


16643.0 


5378.9 


333.6 


11761.7 


3876.0 


341.7 


13202.2 


4179.9 


400.2 


14643.1 


4844.1 


435.5 


18123.4 


5540.5 


423.2 


15553.3 


4992.5 


459.4 


17364.7 


5500.8 


237.5 


12688.3 


3494. 1 


244.8 


11663.0 


3363.4 


236.3 


13435.9 


3533.2 



This disturbance is reflected in the relatively unstable breeding avifaunas of 
Main, North, and South McKittrick canyons. North McKittrick is more 
susceptible to severe flooding than is South McKittrick. The breeding 
avifauna of North McKittrick was the least stable of all. 

Percentage contributions by foraging categories of total biomass varied 
only slightly from year to year in each of the communities studied. Foliage 
insect and ground seed (including fruit) foragers contributed heavily to the 
avifaunal composition of all five communities. The more open vegetative 
canopy of the canyon systems favored more utilization by aerial insect 
feeders than did the more closed canopy of The Bowl. 

Seven foraging classes were represented in Main McKittrick Canyon (Fig. 
3). The increase shown in the foliage insect class for 1974 was due primarily 
to an increase in the breeding population of Black-headed Grosbeaks. 
Yearly fluctuations in the ground insect class resulted from a decrease in 
Bewick's Wrens and House Wrens in 1973 and the utilization of this habitat 
by Hermit Thrushes and Rock Wrens in 1974. The percentage contribution 



222 NEWMAN 

of the foliage seed class was reduced to zero in 1974 because of the absence of 
Scrub Jays. The Ladder-backed Woodpecker did not utilize the Main 
McKittrick study area in 1974, thereby reducing the percentage contribu- 
tion of the timber drilling class to zero for that year. The foraging categories 
timber searching and ground predator were not represented on the study 
plot in Main McKittrick Canyon. 

Six foraging classes were represented in North McKittrick Canyon (Fig. 
5). The foliage seed category, represented by Scrub Jays in Main McKittrick 
Canyon, was absent in this community. Percentage contributions by 
foraging classes remained fairly constant throughout the 3-year study 
period, even though there was a reduced gross avifaunal biomass in North 
McKittrick Canyon in 1973 (Table 8). The reduced total biomass could 
therefore be attributed to a reduction during 1973 in several food sources, 
both vegetative and insect. 

Seven foraging classes are represented in South McKittrick Canyon (Fig. 
7). The timber-searching category, absent in Main and North McKittrick, 
was represented by White-breasted Nuthatches in this community. Foliage 
insects remain the most important source of food supporting the avifauna in 
this area. The percentage contribution of this category to consuming bio- 
mass increased from 39 percent in 1972 to 47 percent in 1973. Nearly all bird 
populations belonging to this foraging category increased in 1973 with the 
exception of Solitary Vireos (Table 10). The percentage contribution of the 
ground-insect class dropped in consuming biomass from 9 percent in 1972 to 
a low of 4 percent in 1974. This change was due primarily to a decreased 
utilization of the study area by Canyon Wrens. White-breasted Nuthatches 
did not nest on the study plot in 1973, thereby reducing the percentage con- 
tribution of the timber-searching category to zero for that year. 

Eight foraging classes were represented each year in Upper Dog Canyon 
(Fig. 9). The increase in percentage contribution from 1972 to 1973 shown in 
the foliage-insect category was influenced by increases in Solitary Vireos, 
Blue-gray Gnatcatchers, and Virginia's Warblers (Table 12). A further 
increase in 1974 can be attributed to population increases in Warbling 
Vireos, Grace's Warblers, Bushtits, Yellow-rumped Warblers, and Scott's 
Orioles. An increase in the percentage contribution of the ground-seed 
category in 1973 resulted from increases in Brown-headed Cowbirds and 
Rufous-crowned Sparrows. The drop in percentage contribution in the 
foliage-seed category from 1972 to 1973 is the result of reduced utilization by 
Acorn Woodpeckers and an absence of Steller's and Scrub Jays in 1973. 
Scrub Jays once again utilized this study area in 1974. 

Nine foraging classes were represented in The Bowl (Fig. 11). The 
ground-predator category was represented by Saw-whet Owls which 
utilized this study area in 1973. Aerial insects as an avifaunal food source 
contributed considerably less to total consuming biomass in The Bowl than 
they did in the other study areas. The increase shown in this category from 
1972 to 1973 resulted in part from increases in populations of Yellow- 
rurnpeu Warblers, Western Tanagers, Orange-crowned Warblers, and 



BREEDING AVIFAUNAS 



223 



TABLE 17. Coefficients of similarity of breeding avifaunas between Main McKittrick, North 
McKittrick, South McKittrick, Upper Dog, and The Bowl for 1972, 1973, and 1974. 





Main 


North 


South 






Study area 


McKittrick 


McKittrick 


McKittrick 


Upper Dog 


The Bowl 






1972 








Main McKittrick 


1.00 


0.66 


0.67 


0.50 


0.17 


North McKittrick 


0.66 


1.00 


0.73 


0.54 


0.19 


South McKittrick 


0.67 


0.73 


1.00 


0.64 


0.29 


Upper Dog 


0.50 


0.54 


0.64 


1.00 


0.39 


The Bowl 


0.17 


0.19 

1973 


0.29 


0.39 


1.00 


Main McKittrick 


1.00 


0.71 


0.62 


0.52 


0.16 


North McKittrick 


0.71 


1.00 


0.66 


0.52 


0.19 


South McKittrick 


0.62 


0.66 


1.00 


0.69 


0.30 


Upper Dog 


0.52 


0.52 


0.69 


1.00 


0.34 


The Bowl 


0.16 


0.19 
1974 


0.30 


0.34 


1.00 


Main McKittrick 


1.00 


0.66 


0.60 


0.55 


0.24 


North McKittrick 


0.66 


1.00 


0.65 


0.53 


0.17 


South McKittrick 


0.60 


0.65 


1.00 


0.72 


0.33 


Upper Dog 


0.55 


0.53 


0.72 


1.00 


0.35 


The Bowl 


0.24 


0.17 


0.33 


0.35 


1.00 



Violet-green Swallows. A reduction in the Common Flicker population was 
in part responsible for the decrease in percentage contribution shown for the 
ground-insect category in 1973. Red Crossbills accounted for the higher per- 
centage shown in the foliage-seed category for 1972, and the addition of 
Acorn Woodpeckers in 1974 accounted for the increase of that year. Fluctu- 
ations in the timber-drilling category were caused by reduced numbers of 
Hairy Woodpeckers, the only member of this category, in 1973. 

A comparison of the avifaunal communities using the coefficient of simi- 
larity as described by Beals (1960) shows, as would be expected, that the can- 
yon systems were much more similar to each other than they were to The 
Bowl (Table 17). Coefficient of similarity values ranged from zero, for those 
avifaunal communities least alike, to 1.00 for identical communities. The 
most avifaunally distinct communities were those at the elevation extremes 
(Main McKittrick as compared to The Bowl). Upper Dog Canyon and 
South McKittrick Canyon had relatively similar avifaunas, as did the con- 
tiguous canyons of the McKittrick system. 

Comparable data to that collected for The Bowl were provided for 
coniferous forest regions of Arizona by Carothers et al. (1973). They found 
in a ponderosa pine community of the San Francisco Mountains a breeding 
avifauna consisting of 23 species, 232 pairs per 100 acres, 15,993 g per 100 



224 NEWMAN 

acres standing crop biomass, 3835.9 g per 100 acres consuming biomass, and 
species diversity (Shannon-Weiner index) equal to 2.83. The Arizona study 
also included a fir, pine, aspen community and gave the following results: 27 
species; 253 pairs per 100 acres; 16,011 g per 100 acres standing crop bio- 
mass; 3936.5 g per 100 acres consuming biomass; species diversity equal to 
2.92. The Guadalupe Bowl has 18 and 19 breeding species, respectively, in 
common with the ponderosa pine and fir, pine, aspen communities of the 
San Francisco Mountains. Species diversity in The Bowl was slightly higher 
than that shown for either community of the Arizona study. In pairs per 100 
acres The Bowl was intermediate between the two Arizona sites. Standing 
crop biomass and consuming biomass were less in The Bowl than in either 
Arizona community. The size and sources of the statistical samples used to 
determine bird weights may account for discrepancies in biomass computa- 
tions between the two geographical locations. 

Snyder (1950) reported on the breeding avifaunas of three different 
coniferous plant communities in the mountains of Colorado. Population 
densities per 100 acres were 102 pairs in Douglas fir-ponderosa pine, 94 pairs 
in Engelmann spruce-subalpine fir, and 59 pairs in lodgepoie pine. More 
than twice as many breeding pairs per 100 acres occurred in the coniferous 
forest region of The Bowl than in the Douglas fir-ponderosa pine commu- 
nity of Colorado. 

Wauer (1971) reported 42 breeding species in the deciduous woodland 
communities of the Chisos Mountains. The combined number of breeding 
species for comparable deciduous woodland associations (McKittrick Can- 
yon system and Upper Dog Canyon) of the Guadalupe Mountains was 45. 

Tatschl (1967) found 31 breeding species with a density of 425 breeding 
pairs per 100 acres in the ponderosa pine community of the Sandia Moun- 
tains of New Mexico. This compares to 32 breeding species and 245 breeding 
pairs per 100 acres in the Guadalupe Bowl. 

Salt (1957) stated that the efficiency of an avifauna in terms of energy use 
is indicated by the ratio of consuming biomass to standing crop biomass. 
The smaller the value of the ratio, the more efficient is the avifauna. Salt's 
premise is based on the fact that large birds require less food per gram of 
body weight than do small birds. Of the Guadalupe communities studied, 
The Bowl had the lowest consuming biomass to standing crop biomass 
(CB/ SCB) ratio (0.28 in 1972, 0.29 in 1973, 0.26 in 1974). The CB/ SCB ratio 
of the canyon woodlands ranged from 0.30 to 0.33. Carothers et al. (1973) 
found CB/SCB ratios of 0.25 and 0.24, respectively, in the fir, pine, aspen 
community and in the ponderosa pine community of the San Francisco 
Mountains in Arizona. 

The work of Burleigh and Lowery (1940) provides a document of interest 
for comparison to the present study. Though their paper does not present 
quantitative data on the breeding avifaunas of the Guadalupes, indications 
of abundance are alluded to in the narratives of their species list. Direct com- 
parison on a species basis to their work is presented in the appendix to this 
paper. 



BREEDING AVIFAUNAS 225 

APPENDIX I 

Annotated List of Bird Specimens Collected by 
George A. Newman from the Guadalupe Mountains, Texas 

Listed in this appendix are all of the bird specimens collected from the 
Guadalupe Mountains by George A. Newman from 1969 to 1974. The 
nomenclature used follows that of The A. O. U. Checklist of North American 
Birds (1957), and published supplements to that work. 

Where there has been a significant change in status of a species when com- 
pared to earlier ornithological investigations of the Guadalupe Mountains, 
notation of such change is made. References to Burleigh and Lowery refer to 
their 1940 publication as given in the Literature Cited section of the main 
text. Detailed information on density and biomass of bird species which 
breed in the woodland areas of the southern Guadalupe Mountains may be 
found in the text of this report and is not repeated here. 

There are 302 specimens representing 99 species and 26 families making up 
this collection. The numbers given after the initials GAN are the field 
numbers of the author. Deposition of the skins is indicated by AMNH 
(American Museum of Natural History, New York), A&M (Texas Coopera- 
tive Wildlife Collection — Texas A&M University, College Station, Texas), 
H-SU (Hardin-Simmons University, Abilene, Texas). 



Accipiter cooperii (Bonaparte), Cooper's Hawk 
1972.— 1 adult female, Upper Dog Canyon, 19 June, GAN #302, A&M. 
1973.— 1 adult female, The Bowl, 26 June, GAN #439, H-SU. 

This species appears to be more common now than when Burleigh and Lowery did their 
research. I found active nests in Upper Dog Canyon and in Smith Canyon. 

Buteo jamaicensis calurus (Cassin), Red-tailed Hawk 
1974.— 1 imm. male, The Bowl, 16 June, GAN #499, A&M. 

The specimen collected was acquiring its first adult plumage; it shows one almost full length 
red rectrix along with the typically barred rectrices of immature plumage. 

This species regularly breeds in the region of The Bowl and along the high cliffs of Upper Dog 
Canyon. 

Actitis macularia (Linnaeus), Spotted Sandpiper 
1972.— 1 adult male, Main McKittrick Canyon, 31 May, GAN #272, A&M. 
This species was seen only rarely along the stream in McKittrick Canyon. 
Columba fasciata fasciata (Say), Band-tailed Pigeon 
1970.— 1 adult female, South McKittrick Canyon, 21 May, GAN #185, H-SU. 

This species was found to be fairly common as a regular breeding bird along the wooded 
canyon slopes of upper South McKittrick. It was also seen regularly during the early summer in 
Upper Dog Canyon and in The Bowl. Burleigh and Lowery found this species to be fairly rare. 

Otus flammeolus flammeolus (Kaup), Flammulated Owl 
1973.— 1 adult male, The Bowl, 25 June, GAN #435, H-SU. 
This secretive species was found regularly in The Bowl, but in very few numbers. 
Bubo virginianus pallescens (Stone), Great Horned Owl 
1973.— 1 adult female, Upper Dog Canyon, 11 June, GAN #408, A&M. 

This species is a regular breeding bird in the Upper Dog Canyon area. I have also heard it 
calling and have seen it in McKittrick Canyon and The Bowl in early summer. 



226 NEWMAN 

Strix occidentalis lucida (Nelson), Spotted Owl 
1972.— 1 adult female, Smith Canyon, 30 December, GAN #364, H-SU. 

This species was also recorded in Upper Dog Canyon during early summer 1974. It is of rare 
occurrence as a breeding species in the Guadalupes. 

Aegolius acadicus acadicus (Gmelin), Saw-whet Owl 
1973.— 1 imm. male, The Bowl, 23 June, GAN #433, A&M; 1 imm. female, The Bowl, 23 June, 

GAN #434, H-SU. 

This species was not recorded by Burleigh and Lowery. My only sighting was of five imma- 
ture birds which responded to my call during the early evening of 23 June 1973, while I was 
camped in The Bowl. 

Caprimulgus vociferus arizonae (Brewster), Whip-poor-will 
1969.— 1 adult male, The Bowl, 3 June, GAN #149, AMNH. 
1973.— 1 adult male, The Bowl, 26 June, GAN #440, H-SU. 

I found this species to be fairly common as a breeding bird in The Bowl; recorded by Burleigh 
and Lowery as "rather scarce." 

Phalaenoptilus nuttallii nuttallii (Audubon), Poor-will 
1974.— 1 adult female, The Bowl, 17 June, GAN #500, H-SU. 

This species was found to be common in the breeding season along the mountain slopes of 
the major canyon systems and in The Bowl. 

Chordeiles minor howellii (Oberholser), Common Nighthawk 
1969.— 1 adult female, The Bowl, 4 June, GAN #150, AMNH. 
1973.— 1 adult female, The Bowl, 25 June, GAN #438, A&M. 

This species was seen commonly each summer (1972, 1973, 1974) feeding at dusk over the 
meadow in The Bowl. Verification of subspecific identification was made by John Hubbard. 

Aeronautes saxatalis saxatalis (Woodhouse), White-throated Swift 
1970.— 1 adult female, South McKittrick Canyon, 22 May, GAN #204, H-SU. 

Archilochus alexandri (Bourcier and Mulsant), Black-chinned Hummingbird 
1970.— 1 adult male, North McKittrick Canyon, 22 May, GAN #207, H-SU; 1 adult female, 

South McKittrick Canyon, 21 May, GAN #208, AMNH. 
1973.— 1 adult male, South McKittrick Canyon, 30 May, GAN #397, A&M; 1 adult female, 

South McKittrick Canyon, 29 May, GAN #395, H-SU; 1 adult female, Upper Dog Canyon, 

12 June, GAN #411, H-SU. 
1974.— 1 adult female, South McKittrick Canyon, 28 May, GAN #475, H-SU. 

This species was found by Burleigh and Lowery to "breed rather sparingly" and was recorded 
by them only in Pine Springs Canyon and Guadalupe Canyon. I found this species to be com- 
mon as a breeding bird in the McKittrick Canyon system and in Upper Dog Canyon. 

Selasphorus platycercus platycercus (Swainson), Broad-tailed Hummingbird 
1971.— 1 adult male, Smith Canyon, 24 May, GAN #231, AMNH. 
1974.— 1 adult male, South McKittrick Canyon, 28 May, GAN #476, H-SU. 

This species regularly breeds in the McKittrick Canyon system, Upper Dog Canyon, and The 
Bowl. 

Eugenes fulgens aureoviridis (van Rossem), Rivoli's Hummingbird 
1972.— 1 adult female, South McKittrick Canyon, 6 June, GAN #280, H-SU. 
1974.— 1 adult female, North McKittrick Canyon, 2 June, GAN #480, A&M. 

This species was not recorded by Burleigh and Lowery. Though rare, this species was seen 
fairly regularly in the McKittrick Canyon system and/ or Upper Dog Canyon in the summers of 
1972, 1973, 1974. 

Colaptes auratus cafer (Gmelin), Common Flicker 
1969.— 1 adult male, The Bowl, 6 June, GAN #151, H-SU; 1 adult female, The Bowl, 6 June, 

GAN #153, H-SU. 
1970.— 1 adult male, The Bowl, 2 June, GAN #216, AMNH. 
1972.— 1 adult female, Upper Dog Canyon, 2 1 June, GAN #316, H-SU; 1 adult male, the Bowl, 

8 July, GAN #340, A&M. 
1973.— 1 adult female, Frijole, 4 January, GAN #383, A&M. 



BREEDING AVIFAUNAS 227 

Colaptes auratus cafer * auratus, Common Flicker 
1969.— 1 adult female, The Bowl, 4 June, GAN #154, H-SU; 1 adult male, The Bowl, 5 June, 
GAN#152, A&M. 

Melanerpes formicivorus formicivorus (Swainson), Acorn Woodpecker 
1969.— 1 adult female, The Bowl, 7 June, GAN #155, H-SU. 
1973.— 1 adult male, Upper Dog Canyon, 13 June, GAN #415, A&M. 

Burleigh and Lowery recorded this species only in The Bowl. I found it to be fairly common as 
a breeding bird in Upper Dog Canyon and of rather spasmodic occurrence in The Bowl. 

Sphyrapicus varius nuchalis (Baird), Yellow-bellied Sapsucker 
1971.— 1 adult female, The Bowl, 3 June, GAN #240, AMNH. 
1973.— 1 adult female, South McKittrick Canyon, 12 October, GAN #445, A&M. 

This species was found only in October by Burleigh and Lowery. The June specimen was 
probably a late transient rather than a breeding bird. The October 1973 specimen was identified 
by John Hubbard as S. v. nuchalis, but as showing some affinity for S. v. varius. 

Sphyrapicus varius varius (Linnaeus), Yellow-bellied Sapsucker 
1973.— 1 adult female, South McKittrick Canyon, 5 January, GAN #385, H-SU. 
Verification of subspecific identification was made by John Hubbard. 

Dendrocopos villosus leucothorectis (Oberholser), Hairy Woodpecker 
1970.— 1 adult male, The Bowl, 2 June, GAN #215, AMNH. 

1972.— 1 adult female, Upper Dog Canyon, 22 June, GAN #3 18, A&M; 1 adult male, The Bowl, 
10 July, GAN #342, H-SU. 
Verification of subspecific identification was made by John Hubbard. 

Dendrocopos scalaris symplectus (Oberholser), Ladder-backed Woodpecker 
1971.— 1 adult male, South McKittrick Canyon, 28 May, GAN #234, AMNH. 
1973.— 1 adult female, Frijole, 4 January, GAN #378, H-SU. 

Both of these specimens were identified as D. s. symplectus by John Hubbard, who states that 
they do show some overlap with D. s. cactophilus. 

Dendrocopos scalaris ssp., Ladder-backed Woodpecker 
1972.— 1 adult male, Smith Canyon, 14 June, GAN #290, H-SU. 
1973.— 1 adult male, Upper Dog Canyon, 14 June, GAN #420, A&M. 
Subspecific identification uncertain at this time. 

Tyrannus vociferans vociferans (Swainson), Cassin's Kingbird 
1971.— 1 adult male, Main McKittrick Canyon, 27 May, GAN #238, H-SU; 1 adult female, 

Main McKittrick Canyon, 27 May, GAN #239, H-SU. 
1972.— 1 adult male, Upper Dog Canyon, 23 June, GAN #324, A&M. 
1973.— 1 adult female, Frijole, 6 June, GAN #404, H-SU. 

Myiarchus cinerascens cinerascens (Lawrence), Ash-throated Flycatcher 
1970.— 1 adult male, South McKittrick Canyon, 21 May, GAN #194, AMNH; 1 adult male, 

North McKittrick Canyon, 22 May, GAN #195, H-SU. 
1971.— 1 adult male, North McKittrick Canyon, 26 May, GAN #236, H-SU. 
1972.— 1 adult female, Upper Dog Canyon, 19 June, GAN #305, A&M. 
Sayornis nigricans semiatra (Vigors), Black Phoebe 
1972.— 1 imm., The Bowl, 9 July, GAN #341, H-SU. 

My only sighting of this species in the Guadalupe Mountains was of the immature individual 
which I collected. Burleigh and Lowery did not record this species. 

Sayornis saya saya (Bonaparte), Say's Phoebe 
1972.— 1 adult male, Upper Dog Canyon, 23 June, GAN #323, H-SU. 
1973.— 1 adult female, Upper Dog Canyon, 15 June, GAN #421, H-SU. 
1974.— 1 adult male, Upper Dog Canyon, 6 June, GAN #482, A&M. 

Empidonax oberholseri (Phillips), Dusky Flycatcher 
1972.— 1 adult female, Upper Dog Canyon, 20 June, GAN #308, H-SU; 1 adult female, Upper 

Dog Canyon, 24 June, GAN #327, H-SU. 
1974.— 1 adult female, Main McKittrick Canyon, 25 May, GAN #470, A&M; 1 adult female, 
Upper Dog Canyon, 7 June, GAN #483, A&M. 



228 NEWMAN 



The breeding status of this species is yet uncertain in the Guadalupe Mountains. Verification 
of species identification was made by John Hubbard. 

Empidonax difficilis difficilis (Baird), Western Flycatcher 
1969.— 1 adult female, The Bowl, 28 May, GAN #159, H-SU. 
1970.— 1 adult male, South McKittrick Canyon, 27 May, GAN #192, AMNH. 

Verification of subspecific identification was made by John Hubbard. 

Empidonax difficilis hellmayri (Brodkorb), Western Flycatcher 
1969.— 1 adult male, The Bowl, 29 May, GAN #158, H-SU. 
1970.— 1 adult male, South McKittrick Canyon, 28 May, GAN #193, AMNH. 
1972.— 1 adult male, The Bowl, 10 July, GAN #344, H-SU. 
1973.— 1 adult male, South McKittrick Canyon, 30 May, GAN #398, H-SU; 1 adult male, 

Upper Dog Canyon, 18 June, GAN #429, A&M. 
1974.— 1 adult male, Upper Dog Canyon, 22 May, GAN #461, H-SU. 

Contopus sordidulus veliei (Coues), Western Wood Pewee 
1969.— 1 adult female, The Bowl, 3 June, GAN #157, H-SU. 
1970.— 2 adult males, South McKittrick Canyon, 21 May, GAN #190, #189, AMNH; 1 adult 

male, North McKittrick Canyon, 22 May, GAN #191, H-SU. 
1971.— 1 adult female, North McKittrick Canyon, 26 May, GAN #230, H-SU. 
1972.— 1 adult female, Guadalupe Spring, 16 June, GAN #294, H-SU; 1 adult male, Upper Dog 

Canyon, 23 June, GAN #325, H-SU. 
1973.— 1 adult male, Upper Dog Canyon, 18 June, GAN #428, A&M. 

Nuttallornis borealis (Swainson), Olive-sided Flycatcher 
1971.— 1 adult female, Smith Canyon, 24 May, GAN #235, AMNH. 
1973.— 1 adult female, Main McKittrick Canyon, 31 May, GAN #399, H-SU. 
1974.— 1 adult female, Main McKittrick Canyon, 25 May, GAN #467, A&M. 

This species was recorded as a breeding bird in the Bowl by Burleigh and Lowery. I did not 
find it in The Bowl, but I did record it as a possible breeding species in Main McKittrick 
Canyon. 

Tachycineta thalassina lepida (Mearns), Violet -green Swallow 
1969.— 1 adult female, The Bowl, 7 June, GAN #156, H-SU. 
1970.— 1 adult female, 1 adult male, The Bowl, 2 June, GAN #206, #205, AMNH. 
1971.— 1 adult male, Main McKittrick Canyon, 27 May, GAN #232, A&M. 
1972.— 1 adult male, Upper Dog Canyon, 22 June, GAN #320, H-SU. 

Listed by Burleigh and Lowery as occurring in summer at an altitude above 7000 ft. I found 
this species to be fairly common during the summer in the wooded canyons and at Manzanita 
Spring at elevations down to 5200 ft. 

Cyanocitta stelleri macrolopha (Baird), Steller's Jay 
1969.— 1 adult male, The Bowl, 27 May, GAN #160, H-SU. 
1970.— 1 adult male, The Bowl, 1 June, GAN #184, AMNH. 
1972.— 1 imm., Upper Dog Canyon, 25 June, GAN #330, H-SU. 
1973.— 1 adult female, Frijole, 1 January, GAN #367, A&M. 

Aphelocoma coerulescens woodhouseii (Baird), Scrub Jay 
1972.— 1 imm. female, Upper Dog Canyon, 20 June, GAN #312, H-SU; 1 imm. male, Upper 

Dog Canyon, 26 June, GAN #331, A&M. 
1973.— 1 adult female, Frijole, 3 January, GAN #376, H-SU. 
1974.— 1 adult male, Upper Dog Canyon, 10 June, GAN #496, H-SU. 

Placement of these specimens in this subspecific category is tentative at this time. 
Nucifraga columbiana (Wilson), Clark's Nutcracker 
1969.— 1 adult male, The Bowl, 1 June, GAN #162, H-SU; 1 adult female, The Bowl, 1 June, 

GAN #161, A&M. 

This species was not recorded by Burleigh and Lowery. A flock of seven nutcrackers, two of 
which were collected, were seen regularly from 31 May to 7 June 1969, while 1 was camped in 
The Bowl. 



BREEDING AVIFAUNAS 229 



Par us gambeli gambeli (Ridgway), Mountain Chickadee 
1969.— 1 adult male, The Bowl, 28 May, GAN #163, H-SU. 
1970.— 1 adult male, The Bowl, 2 June, GAN #210, AMNH. 
1972.— 1 adult male, Upper Dog Canyon, 20 June, GAN #310, A&M. 
1973.— 1 adult male, South McKittrick Canyon, 28 May, GAN #393, H-SU; 1 adult female, 

Main McKittrick Canyon, 5 January, GAN #384, H-SU. 

This species breeds sparingly in wooded canyons down to an elevation of about 5200 ft; this is 
a common nesting bird of The Bowl at 8000 ft elevation. 

Parus inornatus ridgwayi (Richmond), Plain Titmouse 
1972.— 1 adult female, Frijole, 29 December, GAN #355, H-SU. 
1973.— 1 adult female, Frijole, 6 June, GAN #405, A&M. 

This species puzzled Burleigh and Lowery because it seemed to be absent during the breeding 
season. I found a pair, with young, that were using a dead pear tree as a nesting site near the 
Frijole Ranger Station in the summer of 1973. Verification of subspecific identification was 
made by John Hubbard. 

Psaltriparus minimus plumbeous (Baird), Bushtit 
1970.— 1 adult male, South McKittrick Canyon, 28 May, GAN #211, AMNH. 
1971.— 1 adult male, South McKittrick Canyon, 25 May, GAN #267, H-SU. 
1972.— 1 adult (sex undetermined), Main McKittrick Canyon, 8 June, GAN #283, H-SU. 
1973. — 1 adult male, Upper Dog Canyon, 15 June, GAN #423, A&M; 1 imm., Upper Dog 

Canyon, 15 June, GAN #422, A&M. 

Sitta carolinensis nelsoni (Mearns), White-breasted Nuthatch 
1969.— 1 adult female, The Bowl, 5 June, GAN #164, H-SU. 
1972.— 1 adult male, South McKittrick Canyon, 31 May, GAN #274, H-SU; 1 adult male, 

Upper Dog Canyon, 21 June, GAN #317, A&M. 

I found this species, as did Burleigh and Lowery, to be a common breeding bird in The Bowl. 
It is also commonly found breeding in Upper Dog Canyon. 

Sitta canadensis (Linnaeus), Red-breasted Nuthatch 
1973.— 1 adult (sex undetermined), The Bowl, 13 October, GAN #446, A&M. 

Sitta pygmaea melanotis (van Rossem), Pygmy Nuthatch 
1969.— 1 adult male, The Bowl, 3 June, GAN #165, H-SU. 
1970.— 1 adult male, The Bowl, 1 June, GAN #196, AMNH. 
1972.— 1 imm., Upper Dog Canyon, 20 June, GAN #315, A&M. 
1974.— 1 imm. male, The Bowl, 18 June, GAN #501, H-SU. 

Breeds commonly in The Bowl and is a sporadic breeder in Upper Dog Canyon. 
Certhia familiar is montana (Ridgway), Brown Creeper 
1969.— 1 adult female, The Bowl, 29 May, GAN #166, H-SU. 
1972.— 1 imm., The Bowl, 8 July, GAN #339, A&M. 
1973.— 1 adult male, Main McKittrick Canyon, 5 January, GAN #386, H-SU. 

Subspecific identification was verified by John Hubbard. 

Troglodytes aedon parkmanii (Audubon), House Wren 
1969.— 1 adult female, The Bowl, 29 May, GAN #168, H-SU. 
1972.— 1 imm., The Bowl, 8 July, GAN #338, A&M. 

Appears to have decreased in abundance since Burleigh and Lowery researched the 
Guadalupes. 

Thryomanes bewickii eremophilus (Oberholser), Bewick's Wren 
1970.— 1 adult female, North McKittrick Canyon, 22 May, GAN #212, AMNH. 
1972.— 1 adult male, Upper Dog Canyon, 27 June, GAN #333, H-SU. 
1973.— 1 adult male, South McKittrick Canyon, 2 June, GAN #401, A&M. 

Campy lor hynchus brunneicapillus couesi (Sharpe), Cactus Wren 
1973.— 1 adult male, Frijole, 2 January, GAN #374, H-SU. 
1974.— 1 imm., PX Flat, 9 June, GAN #491, A&M. 

This species was thought by Burleigh and Lowery not to breed at an elevation higher than 



230 NEWMAN 

5600 ft in the Guadalupes. I have found it to be a fairly common nesting species in PX Flat at an 
elevation of about 6500 ft. 

Catherpes mexicanus conspersus (Ridgway), Canyon Wren 
1970.— 1 adult male, South McKittrick Canyon, 28 May, GAN #213, AMNH. 
1972.— 1 imm., Upper Dog Canyon, 24 June, GAN #329, A&M. 
1973.— 1 adult female, Frijole, 3 January, GAN #375, H-SU. 

Salpinctes obsoletus obsoletus (Say), Rock Wren 
1969.— 1 adult male, Frijole, 4 April, GAN #167, H-SU. 
1972. — 1 imm. male, Upper Guadalupe Spring, 16 June, GAN #295, H-SU. 
1973. — 1 imm. male, Upper Dog Canyon, 13 June, GAN #413, A&M; 1 adult female, Upper 
Dog Canyon, 13 June, GAN #414, H-SU. 

Mimus polyglottos leucopterus (Vigors), Mockingbird 
1972.— 1 adult male, Lower Guadalupe Spring, 16 June, GAN #296, H-SU. 
1974.— 1 adult male, Upper Dog Canyon, 8 June, GAN #490, A&M. 

Toxostoma curvirostre celsum (Moore), Curve-billed Thrasher 
1972.— 1 adult female, Upper Dog Canyon, 28 June, GAN #334, A&M. 
1973. — 1 adult female, Frijole, 2 January, GAN #372, H-SU; 1 adult female, Upper Dog 
Canyon, 14 June, GAN #419, H-SU. 

Oreoscoptes montanus (Townsend), Sage Thrasher 
1973.— 1 adult male, Frijole, 4 January, GAN #379, A&M. 

Turdus migratorius propinquus (Ridgway), Robin 
1972.— 1 adult male, Frijole, 17 June, GAN #298, A&M; 1 adult female, Smith Canyon, 30 
December, GAN #359, H-SU. 

The status of this species in the Guadalupe Mountains has changed considerably since the 
work of Burleigh and Lowery. They recorded it as a common nesting species of the high 
country. The only Robin I found during the breeding season from 1969 to 1974 was the one 
collected at Frijole. There seems to have been a great decline in the breeding population of this 
species in the Guadalupes. 

Catharus guttatus guttatus (Pallas), Hermit Thrush 
1973. — 1 adult male, Pine Springs Canyon, 1 January, GAN #366, A&M. 

Catharus guttatus slevini (Grinnell), Hermit Thrush 
1973.— 1 adult (sex undetermined), The Bowl, 13 October, GAN #448, H-SU. 
1974.— 1 adult male, Main McKittrick Canyon, 19 March, GAN #455, A&M; 2 adult females, 
Main McKittrick Canyon, 19 March, GAN #456, #457, H-SU. 

Catharus guttatus auduboni (Baird), Hermit Thrush 
1972.— 1 adult female, Upper Dog Canyon, 24 June, GAN #328, H-SU; 1 imm. male, The 

Bowl, 10 July, GAN #343, A&M. 
1974.— 1 adult female, Upper Dog Canyon, 22 May, GAN #459, H-SU. 

Burleigh and Lowery did not find this species occurring as a breeding bird below 8000 ft 
elevation. I found it to occur fairly commonly in Upper Dog Canyon at 6200 ft, as well as in The 
Bowl. Subspecific identification was verified by John Hubbard. 

Catharus ustulatus ustulatus (Nuttall), Swainson's Thrush 
1971.— 1 adult male, South McKittrick Canyon, 25 May, GAN #237, A&M. 

This species was not recorded by Burleigh and Lowery. Though rare, I found it to be fairly 
regular in late spring in the McKittrick Canyon system. Subspecific identification was verified 
by John Hubbard. 

Catharus ustulatus swainsoni (Tschudi), Swainson's Thrush 
1973.— 1 adult male, South McKittrick Canyon, 28 May, GAN #392, H-SU. 
1974.— 1 adult male, Main McKittrick Canyon, 25 May, GAN #471, A&M. 
Subspecific identification was verified by John Hubbard. 

Sialia mexicana bairdi (Ridgway), Western Bluebird 
1969.— 1 adult male, The Bowl, 6 June, GAN #169, H-SU. 
1970.— 1 adult male, The Bowl, 1 June, GAN #226, AMNH. 



BREEDING AVIFAUNAS 231 

1972.— 1 adult male, Upper Dog Canyon, 18 June, GAN #299, A&M; 1 adult male, Frijole, 29 
December, GAN #354, H-SU. 

This species breeds sparingly in The Bowl, but is fairly common during the breeding season in 
Upper Dog Canyon. 

Myadestes townsendi townsendi (Audubon), Townsend's Solitaire 
1972.— 1 adult female, Smith Canyon, 30 December, GAN #361, A&M. 

Polioptila caerulea amoenissima (Grinnell), Blue-gray Gnatcatcher 
1970.— 1 adult male, North McKittrick Canyon, 22 May, GAN #209, AMNH. 
1972.— 1 adult female, North McKittrick Canyon, 3 June, GAN #279, H-SU; 1 imm. male, 
Main McKittrick Canyon, 9 June, GAN #285, K-SU; 1 adult male, Upper Dog Canyon, 24 
June, GAN #326, H-SU. 
1973. — 1 adult male, Upper Dog Canyon, 16 June, GAN #427, A&M. 

Regulus calendula calendula (Linnaeus), Ruby-crowned Kinglet 
1973.— 1 adult female, Main McKittrick Canyon, 12 October, GAN #443, H-SU. 

Anthus spinoletta ssp., Water Pipit 
1974.— 1 adult female, North McKittrick Canyon, 2 June, GAN #481, A&M. 

Bombyciila cedrorum (Vieillot), Cedar Waxwing 
1973.— 1 adult female, Frijole, 4 January, GAN #381, H-SU. 

Vireo vicinior (Coues), Gray Vireo 
1972.— 1 adult male, Main McKittrick Canyon, 31 May, GAN #276, A&M. 
1973.— 1 adult male, Main McKittrick Canyon, 29 June, GAN #441, H-SU. 

Vireo solitarius plumbeus (Coues), Solitary Vireo 
1969.— 1 adult female, The Bowl, 3 June, GAN #170, H-SU. 
1970. — 1 adult female, South McKittrick Canyon, 27 May, GAN #203, H-SU; 1 adult male, 

The Bowl, 1 June, GAN #202, AMNH. 
1972.— 1 adult male, South McKittrick Canyon, 31 May, GAN #277, H-SU; 1 adult female, 
North McKittrick Canyon, 7 June, GAN #282, H-SU; 1 adult male, Upper Dog Canyon, 20 
June, GAN #307, A&M. 

Vireo olivaceus (Linnaeus), Red-eyed Vireo 
1970.— 1 adult male, South McKittrick Canyon, 21 May, GAN #214, AMNH. 
1972.— 1 adult male, South McKittrick Canyon, 31 May, GAN #275, H-SU. 

This species was not recorded by Burleigh and Lowery. I found it to occur sporadically in the 
Guadalupes in late spring. 

Vireo gilvus swainsonii (Baird), Warbling Vireo 
1969.— 1 adult female, The Bowl, 3 June, GAN #171, H-SU. 
1970.— 1 adult male, South McKittrick Canyon, 20 May, GAN #201, H-SU; 1 adult male, 

South McKittrick Canyon, 21 May, GAN #200, AMNH. 
1972.— 1 adult female, Upper Dog Canyon, 19 June, GAN #300, A&M. 

Helmitheros vermivorus (Gmelin), Worm-eating Warbler 
1973.— 1 adult male, South McKittrick Canyon, 28 May, GAN #394, A&M. 

This species was not recorded by Burleigh and Lowery. The collected specimen was taken 
from a mist net and is the only record that I have for this species in the Guadalupes. 

Vermivora celata orestera (Oberholser), Orange-crowned Warbler 
1969.— 1 adult male, The Bowl, 30 May, GAN #172, H-SU. 
1970.— 1 adult male, The Bowl, 1 June, GAN #199, AMNH. 
1973.— t adult female, The Bowl, 25 June, GAN #436, H-SU. 
1974.— 1 adult female, Upper Dog Canyon, 21 May, GAN #458, A&M. 
Subspecific identification verified by John Hubbard. 

Vermivora virginiae (Baird), Virginia's Warbler 
1972.— 1 adult female, Upper Dog Canyon, 20 June, GAN #314, A&M; 1 imm., Upper Dog 

Canyon, 22 June, GAN #319, A&M. 
1973.— 1 adult male, Upper Dog Canyon, 12 June, GAN #410, H-SU. 
1974.— 1 adult male, Upper Dog Canyon, 10 June, GAN #494, H-SU. 



232 NEWMAN 



This species was not recorded as a breeding bird by Burleigh and Lowery. I found it to breed 
fairly commonly in Upper Dog Canyon and to breed sporadically in South McKittrick Canyon. 

Dendroica coronata auduboni (Townsend), Yellow-rumped Warbler 
1969.— 1 adult female, The Bowl, 5 June, GAN #173, H-SU. 
1970.— 1 adult male, The Bowl, 1 June, GAN #198, AMNH. 
1973.— 1 adult male, The Bowl, 25 June, GAN #437, A&M; 1 adult female, The Bowl, 13 

October, GAN #447, H-SU. 
1974.— 1 adult female, Upper Dog Canyon, 7 June, GAN #487, H-SU. 

This species was found most commonly in The Bowl, but it also was a sporadic breeding bird 
in Upper Dog Canyon. 

Dendroica nigrescens (Townsend), Black-throated Gray Warbler 
1974.— 1 adult female, Upper Dog Canyon, 7 June, GAN #485, A&M. 

The collected specimen was not in breeding condition. This species recorded only in October 
by Burleigh and Lowery. 

Dendroica graciae graciae (Baird), Grace's Warbler 
1970.— 1 adult male, The Bowl, 2 June, GAN #197, AMNH. 

1972.— 1 adult female, South McKittrick Canyon, 11 June, GAN #286, H-SU; 1 imm., South 
McKittrick Canyon, 11 June, GAN #287, H-SU; 1 adult female, Upper Dog Canyon, 20 
June, GAN #311, H-SU. 
1973.— 1 adult male, Upper Dog Canyon, 1 1 June, GAN #409, A&M. 

Dendroica pensylvanica (Linnaeus), Chestnut-sided Warbler 
1971.— 1 adult female, Frijole, 24 May, GAN #229, A&M. 

The specimen collected is my only record of this species in the Guadalupes. It was not 
recorded by Burleigh and Lowery. 

Oporornis tolmiei tolmiei (Townsend), MacGillivrays Warbler 
1974.— 1 adult female, Upper Dog Canyon, 22 May, GAN #462, A&M. 
Subspecific identification was verified by John Hubbard. 

Wilsonia pusilla pileolata (Pallas), Wilson's Warbler 
1971.— 1 adult male, South McKittrick Canyon, 25 May, GAN #228, AMNH. 
1974.— 1 adult female, Upper Dog Canyon, 22 May, GAN #460, A&M. 
Wilsonia pusilla pusilla (Wilson), Wilson's Warbler 
1974.— 1 adult female, South McKittrick Canyon, 24 May, GAN #463, H-SU; 1 adult female, 
Upper Dog Canyon, 7 June, GAN #484, H-SU; 1 adult male, Upper Dog Canyon, 10 June, 
GAN #495, A&M. 
Subspecific identification was verified by John Hubbard. 

Setophaga ruticilla ruticilla (Linnaeus), American Redstart 
1974.— 1 adult male, Main McKittrick Canyon, 25 May, GAN #469, A&M. 

I recorded this species only in the late spring of 1974 in McKittrick Canyon. It was not 
recorded by Burleigh and Lowery. 

Passer domesticus domesticus (Linnaeus), House Sparrow 
1973.— 2 adult females, Frijole, 7 June, GAN #406, #407, A&M. 

This species is common locally at Frijole and occurs sparingly around the Upper Dog Canyon 
Ranger Station. It was not found away from human habitation. 

Icterus cucullatus nelsoni (Ridgway), Hooded Oriole 
1972.— 1 adult male, Frijole, 14 June, GAN #289, A&M. 
1974.— 1 adult male, Frijole, 15 June, GAN #498, H-SU. 

This species occurs sparingly around Frijole during the breeding season. It was not recorded 
by Burleigh and Lowery. 

Icterus parisorum (Bonaparte), Scott's Oriole 
1970.— 1 adult male, Main McKittrick Canyon, 29 May, GAN #221, H-SU; 1 adult female, 

Main McKittrick Canyon, 29 May, GAN #220, AMNH. 
1972.— 1 adult male, Upper Guadalupe Spring, 16 June, GAN #297, A&M. 
Molothrus ater obscurus (Gmelin), Brown-headed Cowbird 
1972.— 1 aduit male, Main McKittrick Canyon, 31 May, GAN #273, H-SU; 1 adult male, 
Upper Dog Canyon, 26 June, GAN #332, A&M. 



BREEDING AVIFAUNAS 233 

1973.— 1 adult male, Upper Dog Canyon, 14 June, GAN #417, H-SU; 1 adult male, Upper Dog 

Canyon, 16 June, GAN #426, H-SU. 
1974.— 1 adult female, Frijole, 15 June, GAN #497, H-SU. 

This species occurred fairly commonly in the canyons and was quite common around Frijole. 
John Hubbard identified the male specimens as M. ater obscurus > artemisiae.. 

Piranga ludoviciana (Wilson), Western Tanager 
1969.— 1 adult male, The Bowl, 3 June, GAN #174, A&M. 
1970.— 1 adult male, Main McKittrick Canyon, 22 May, GAN #222, AMNH; 1 adult female, 

North McKittrick Canyon, 22 May, GAN #223, H-SU. 
1971.— 1 adult female, Main McKittrick Canyon, 27 May, GAN #233, AMNH. 
1972.— 1 adult female, Upper Dog Canyon, 19 June, GAN #303, H-SU. 

Burleigh and Lowery recorded this species as being limited during the breeding season to an 
elevation of 8000 ft and above. I found this species to breed regularly in the canyons of the 
Guadalupes down to an elevation of 5200 ft. 

Piranga olivacea (Gmelin), Scarlet Tanager 
1974.— 1 adult male, Main McKittrick Canyon, 25 May, GAN #468, A&M. 

My only record of this species is that of the specimen collected. It was not recorded by 
Burleigh and Lowery. 

Piranga flava hepatica (Swainson), Hepatic Tanager 
1970.— 1 adult male, Main McKittrick Canyon, 23 May, GAN #225, AMNH; 1 adult female, 

South McKittrick Canyon, 28 May, GAN #224, AMNH. 
1971.— 1 adult female, Main McKittrick Canyon, 27 May, GAN #242, H-SU. 
1972.— 1 adult male, Upper Dog Canyon, 22 June, GAN #322, A&M. 
1974.— 1 adult male, South McKittrick Canyon, 24 May, GAN #466. H-SU. 

Piranga rubra cooperi (Ridgway), Summer Tanager 
1972.— 1 adult male, Frijole, 15 June, GAN #292, A&M. 

Subspecific identification was verified by John Hubbard. 

Piranga rubra rubra (Linnaeus), Summer Tanager 
1974.— 1 adult male, South McKittrick Canyon, 24 May, GAN #465, H-SU. 

Subspecific identification was verified by John Hubbard. 

Pyrrhuloxia sinuata sinuata (Bonaparte), Pyrrhuloxia 
1973.— 1 adult female, Frijole, 3 January, GAN #377, H-SU. 

Pheucticus melanocephalus melanocephalus (Swainson), 
Black-headed Grosbeak 
1969.— 1 adult male, The Bowl, 6 June, GAN #175, H-SU. 
1970.— 1 adult male, North McKittrick Canyon, 22 May, GAN #219, AMNH; 1 adult female, 

Main McKittrick Canyon, 28 May, GAN #218, AMNH. 
1972.— 1 adult male, Upper Dog Canyon, 20 June, GAN #313, A&M. 

I found this to be a fairly common breeding species in the canyons and high country of the 
Guadalupes. It was considered uncommon by Burleigh and Lowery. 

Guiraca caerulea interfusa (Dwight and Griscom), Blue Grosbeak 
1972.— 1 adult female, Main McKittrick Canyon, 1 June, GAN #278, A&M. 
1974.— 1 adult female, Main McKittrick Canyon, 25 May, GAN #473, H-SU. 

I found this species to occur regularly at the mouth of McKittrick Canyon. 
Passerina cyanea (Linnaeus), Indigo Bunting 
1974.— 1 adult male, South McKittrick Canyon, 27 May, GAN #474, A&M; 1 adult male, Main 

McKittrick Canyon, 29 May, GAN #477, H-SU. 

Several Indigo Buntings were seen near the water in Main and South McKittrick canyons in 
late May 1974. These were my only sightings of this species in the Guadalupes. This species was 
not recorded in the Guadalupes by Burleigh and Lowery. 

Carpodacus mexicanus frontalis (Say), House Finch 
1972.— 1 adult female, Main McKittrick Canyon, 13 June, GAN #288, H-SU; 1 adult male, 

Upper Dog Canyon, 20 June, GAN #309, A&M. 
1973.— 1 adult male, Main McKittrick Canyon, 4 June, GAN #402, H-SU; 1 adult male, Upper 

Dog Canyon, 12 June, GAN #412, H-SU. 



234 NEWMAN 

1974.— 1 adult female, Upper Dog Canyon, 8 June, GAN #489, H-SU. 

Spinus pinus pinus (Wilson), Pine Siskin 
1970.— 1 adult male, The Bowl, 2 June, GAN #186, AMNH. 
1973.— 1 adult female, Frijole, 4 January, GAN #382, H-SU; 1 adult male, South McKittrick 

Canyon, 29 May, GAN #396, A&M. 
1974.— 1 adult female, South McKittrick Canyon, 24 May, GAN #464, H-SU. 

Spinus psaltria psaltria (Say), Lesser Goldfinch 
1971.— 1 adult male, South McKittrick Canyon, 25 May, GAN #227, AMNH. 
1972.— 1 adult female, Main McKittrick Canyon, 8 June, GAN #284, H-SU; 1 adult male, 

Upper Dog Canyon, 19 June, GAN #306, A&M. 
1973.— 1 adult male, Upper Dog Canyon, 13 June, GAN #416, H-SU; 1 adult male, Upper 

Dog Canyon, 14 June, GAN #418, H-SU; 2 adult males, Upper Dog Canyon, 18 June, 

GAN #430, #431, H-SU. 
1974.— 1 adult male, Upper Dog Canyon, 8 June, GAN #488, H-SU. 

This species appears to be much more common than it was when Burleigh and Lowery did 
their work. Subspecific identification was verified by John Hubbard. 

Loxia curvirostra bendirei (Ridgway), Red Crossbill 

1972.— 1 adult male, The Bowl, 7 July, GAN #336, H-SU; 1 adult female, The Bowl, 7 July, 

GAN #337, A&M. 

1974.— 1 adult male, The Bowl, 17 March, GAN #454, H-SU. 

The female specimen of this subspecies was carrying a well-developed egg ( 1 9 mm * 1 3 mm) in 
the oviduct. Identification of subspecies was made by Alan R. Phillips. 
Loxia curvirostra benti (Griscom), Red Crossbill 
1972.— 1 adult female, The Bowl, 7 July, GAN #335, H-SU. 

Subspecific identification by Alan R. Phillips. 

Chlorura chlorura (Audubon), Green-tailed Towhee 
1974.— 1 adult female, South McKittrick Canyon, 30 May, GAN #478, A&M. 

Pipilo erythropthalmus gaigei (Van Tyne and Sutton), Rufous-sided Towhee 
1969.— 1 adult female, The Bowl, 3 June, GAN #177, H-SU. 
1970.— 1 adult male, South McKittrick Canyon, 21 May, GAN #217, AMNH. 
1972.— 1 adult male, North McKittrick Canyon, 7 June, GAN #281, H-SU; 1 imm., Upper 

Dog Canyon, 22 June, GAN #321, A&M. 

Subspecific identification was verified by John Hubbard. 

Pipilo fuscus mesoleucus (Baird), Brown Towhee 
1969.— 1 adult female, Frijole, 4 April, GAN #176, H-SU. 
1972.— 1 adult male, Upper Dog Canyon, 19 June, GAN #304, H-SU; 1 adult male, Frijole, 29 

December, GAN #358, A&M. 
1973.— 1 adult female, Main McKittrick Canyon, 4 June, GAN #403, H-SU. 
Aimophila ruflceps scottii * eremoeca, Rufous-crowned Sparrow 
1971.— 1 adult female, Main McKittrick Canyon, 27 May, GAN #241, AMNH. 
1972.— 1 adult female, Upper Guadalupe Spring, 16 June, GAN #293, H-SU. 
1974.— 1 adult female, Upper Dog Canyon, 7 June, GAN #486, H-SU; 1 adult male, PX Flat, 9 

June, GAN #492, A&M. 

Subspecific identification was made by John Hubbard. 

Amphispiza bilineata opuntia (Burleigh and Lowery), Black-throated Sparrow 
1972.— 1 adult male, Smith Canyon, 14 June, GAN #291, H-SU. 
1974.— 1 adult female, PX Flat, 9 June, GAN #493, A&M. 

Subspecific identification was verified by John Hubbard. 

Junco hyemalis montanus, Dark-eyed Junco 
1972.— 1 adult female, Frijole, 29 December, GAN #357, H-SU. 
1973.— 1 adult female, Frijole, 2 January, GAN #370, H-SU; 1 adult (sex undetermined), 2 

January, GAN #371, H-SU. 

Subspecific identification was made by John Hubbard. 



BREEDING AVIFAUNAS 235 

Junco hyemalis hyemalis (Linnaeus), Dark-eyed Junco 
1973.— 1 adult male, Frijole, 1 January, GAN #368, H-SU. 
Subspecific identification was made by John Hubbard. 

Junco caniceps caniceps (Woodhouse), Gray-headed Junco 
1972.— 1 adult male, Frijole, 30 December, GAN #363, A&M. 
1973.— 1 adult (sex undetermined), Frijole, 2 January, GAN #369, H-SU. 
1974.— 1 adult female, Main McKittrick Canyon, 31 May, GAN #479, H-SU. 
Subspecific identification was made by John Hubbard. 

Junco caniceps dorsalis (Henry), Gray-headed Junco 
1969.— 2 adult females, The Bowl, 28 May, GAN #178, A&M, GAN #179, H-SU. 
1972.— 1 imm. female, The Bowl, 11 July, GAN #345, H-SU. 
Subspecific identification of adult specimens was made by John Hubbard. 
Spizella passerina arizonae (Coues), Chipping Sparrow 
1969.— 1 adult male, The Bowl, 3 June, GAN #180, H-SU. 
1972.— 1 imm., Upper Dog Canyon, 19 June, GAN #301, H-SU; 1 adult male, Frijole, 29 

December, GAN #356, H-SU; 1 adult male, Frijole, 30 December, GAN #360, H-SU. 
1973. — 1 adult male, Upper Dog Canyon, 16 June, GAN #425, H-SU; 1 adult female, Upper 
Dog Canyon, 16 June, GAN #424, A&M; 1 adult male, South McKittrick Canyon, 12 
October, GAN #444, H-SU. 

Spizella atrogularis evura (Coues), Black-chinned Sparrow 
1970.— 1 adult male, Main McKittrick Canyon, 23 May, GAN #188, H-SU; 1 adult male, Main 

McKittrick Canyon, 28 May, GAN #187, A&M. 
1973.— 1 adult male, Frijole, 2 January, GAN #373, H-SU; 1 adult male, Main McKittrick 
Canyon, 1 June, GAN #400, H-SU. 

Zonotrichia leucophrys gambellii (Nuttall), White-crowned Sparrow 
1973.— 1 adult female, Frijole, 4 January, GAN #380, A&M. 
1974.— 1 adult female, Main McKittrick Canyon, 25 May, GAN #472, H-SU. 
Subspecific identification verified by John Hubbard. 

Zonotrichia albicollis (Gmelin), White-throated Sparrow 
1972.— 1 adult male, Frijole, 30 December, GAN #362, H-SU. 

Melospiza lincolnii lincolnii (Audubon), Lincoln's Sparrow 
1972.— 1 adult (sex undetermined), Smith Canyon, 30 December, GAN #365, H-SU. 



LITERATURE CITED 

American Ornithologists' Union. 1957 Checklist of North American birds, 

A.O.U., Baltimore, 5th ed., 691 pp. 
Anderson, S. H. 1970. The avifaunal composition of Oregon white oak stands. 

Condor 72:417-423. 
Bailey, V. 1905. Biological survey of Texas. N. Am. Fauna 25:1-222. 
Beals, E. 1960. Forest birds communities in the Apostle Islands of Wisconsin. 

Wilson Bull. 72:156-181. 
Biaggi, V., Jr. 1960. The birds of Culberson County, Texas, with notes on eco- 
logical aspects. Tex. Ornithol. Soc. Newsletter 8(8): 1-18; 8(10): 1-20. 
Burleigh, T. D., and G. H. Lowery. 1940. Birds of the Guadalupe Mountain 

Region of Western Texas. Occas. Pap. Mus. Zool, La. State Univ. 8:85-151. 
Carothers, S. W., J. R. Haldeman, and R. P. Balda. 1973. Breeding birds of 

the San Francisco Mountain area and the White Mountains, Arizona. Mus. N. 

Ariz. Tech. Ser. 12:1-36. 
Cottam, G., and J. T. Curtis. 1956. The use of distance measures in phyto- 

sociological sampling. Ecology 37:451-460. 



236 NEWMAN 

Cys, J. M. 1971. Discussion: Origin of Capitan formation, Guadalupe Mountains, 
New Mexico and Texas. Am. Assoc. Pet. Geol. Bull. 55:310-312. 

Gehlbach, F. R. 1967. Vegetation of the Guadalupe Escarpment, New Mex- 
ico-Texas. Ecology 48:404-419. 

Hubbard, J. P. 1965. The summer birds of the forests of the Mogollon Moun- 
tains, New Mexico. Condor 67:404-415. 

Johnson, N. K. 1974. Montane avifaunas of southern Nevada: historical change in 
species composition. Condor 76:334-337. 

Karr, J. R. 1968. Habitat and avian diversity on strip-mined land in east-central 
Illinois. Condor 70:348-357. 

Kendeigh, S. C. 1944. Measurement of bird populations. Ecol. Monogr. 14:67- 
106. 

Levy, B. 1971. Guadalupe Mountains National Park: historic resource study. Dep. 
Inter. Div. Hist., Off. Archeol. Hist. Preserv., Washington, D.C., 137 pp. 

Marshall, J. T., Jr. 1957. Birds of pine-oak woodland in southern Arizona and 
adjacent Mexico. Pac. Coast Avifauna 32:1 125. 

Newell, N. D., J. K. Rigby, A. G. Fischer, A. J. Whiteman, J. E. Hickox, 
and J. S. Bradley. 1972. The Permian Reef Complex of the Guadalupe 
Mountains region, Texas and New Mexico: A Study in Paleoecology. Haffner 
Publ. Co., New York, 236 pp. 

Newman, G. A. 1971. Guadalupe. Bull. Tex. Ornithol. Soc. 4:2-4. 

1974a. Recent bird records from the Guadalupe Mountains, Texas. South- 
west. Nat. 19:1-7. 

1974b. Check-list of Birds: Guadalupe Mountains National Park. Pri- 



vately published. 
Salt, G. W. 1953. An ecologic analysis of three California avifaunas. Condor 

55:258-273. 
1957. An analysis of avifaunas in the Teton Mountains and Jackson Hole, 

Wyoming. Condor 59:373-393. 
Snyder, D. P. 1950. Bird communities in the coniferous forest biome. Condor 

52:17-27. 
Tatschl , J. L. 1967. Breeding birds of the Sandia Mountains and their ecological 

distributions. Condor 69:479-490. 
Wauer, R. H. 1971. Ecological distribution of birds of the Chisos Mountains, 

Texas. Southwest. Nat. 16:1-29. 



ACKNOWLEDGMENTS 

I extend appreciation to the following agencies which have financially 
supported my Guadalupe research over the past several years: Carlsbad 
Caverns Natural History Association (1969, 1970); Frank M. Chapman 
Memorial Fund of the American Museum of Natural History (1970, 1971); 
and especially the Welder Wildlife Foundation and its director, the late Dr. 
Clarence Cottam, for granting me a fellowship in 1972, 1973, and 1974 to 
support my doctoral research. I am grateful to the following National Park 



BREEDING AVIFAUNAS 237 

Service personnel who have been a great help with logistical support: Don 
Dayton, Phil Van Cleave, Gary Ahlstrand, John Chapman, Roger Reisch, 
Jane Tate, Marjorie Glass, Francis Schneider, and Norm and Marjorie 
Stephan. Appreciation also is extended to my students from Hardin- 
Simmons University, J. F. Cadenhead and David Dean, for their untiring 
efforts in performing camp duties; to Dr. John Hubbard for graciously veri- 
fying the subspecific identification of many of my Guadalupe specimens; to 
Dr. Allan Phillips for examining the Red Crossbill specimens; to Dr. Fred 
Gehlbach for his initial encouragement for me to investigate the breeding 
avifaunas of the Guadalupes; and particularly to Drs. Keith Arnold, Jim 
Dixon, Jack Inglis, and Fred Smeins for their supervision and support of 
this research. 

A special thanks goes to my wife, Carolyn, and my children, Laura and 
Jason, for their devotion and understanding as I abandoned them for long 
periods of the summer during the course of this research. 



Post-Pleistocene Mammals from Pratt 
Cave and Their Environmental 
Significance 



ERNEST L. LUNDELIUS, JR., University of Texas, Austin 

This report is one of several resulting from a multidisciplinary study of 
Pratt Cave, its deposits, and the archeological and paleontological materials 
derived from it. Only a brief description of the cave, its location, and the 
character of its deposits will be given here, as these subjects will be dealt with 
in more detail elsewhere. The major subject of this paper is the description 
and analysis of the mammal bones found in the cave. 

Pratt Cave is located in the southwest wall of the mouth of McKittrick 
Canyon near the northwest corner of Sec. 12, Block 65, Twp. 1, T and T 
Survey, in the southern Guadalupe Mountains, Culberson County, Texas. It 
is at the base of a 130-ft (39 m) cliff and at the top of a steep talus slope that 
extends to the bottom of the canyon 300 ft (90 m) below. The approximate 
elevation of the cave is 5300 ft (1590 m). 

The cave was formed by solution near the base of the Lamar member of 
the Bell Canyon Formation of Guadalupian (Upper Middle Permian) age. 
The cave probably formed before or during the early stages of the deposi- 
tion of the Ogallala Formation of Pliocene age (Bretz 1949; Horberg 1949). 

The cave is an elongate tunnel oriented N41°E, S41°W (Fig. 1). Its 
entrance is oval with a horizontal diameter of 8 ft (240 cm) and a vertical 
diameter of about 5 ft (150 cm). The cave widens to approximately 9 ft (270 
cm) within 6 ft (180 cm) of the entrance and then narrows rapidly to a 
passage 3 to 4 ft (90 to 120 cm) in width. The roof of the cave rises rapidly 5 ft 
(150 cm) from the entrance to form a chimney more than 10 ft (300 cm) in 
height, centered over the widest part of the cave. Beyond the chimney, the 
roof height varies from 3 to 4 ft (90 to 120 cm). 

The cave fill consists of fine sand and silt, with a concentration of platy 
limestone chips in the lowest part. There is a general gradation upward from 
this zone into subangular to subround grains of limestone. Quartz and 
calcite grains and abundant pebble to cobble-sized rock fragments are found 
throughout the deposit. The latter are plate-like fragments that have spalled 
off the roof and walls. The finer-grained material probably had an aeolian 

239 



240 LUNDELIUS 



19 22 25 28 




SEC D 



unexcovoted 
/ 



SEC A 



scale in fee 



Fig. 1. Plan view of Pratt Cave showing horizontal grid and location of the hearth 
and cache. 



origin. There is no apparent internal stratigraphy, and correlation of dif- 
ferent parts of the cave is difficult. 

The deposits are thin, ranging from 2.5 ft (75 cm) in the front part of the 
cave to nothing toward the back where bedrock is exposed at the surface 
(Fig. 2). Two holes in the center of the cave are evidence of disturbance by 
vandals. In addition to these, an earlier disturbance was found — a hole filled 
with burned rocks and bone fragments, remnants of a hearth. 

Mammal bones from Pratt Cave provide an opportunity to investigate 
postglacial faunal changes in a mountainous region of the southwestern 
United States. Previous work has shown that the regional Pleistocene fauna 
contained, in addition to the extinct species, extant forms such as Marmota 




Fig. 2. Longitudinal section of Pratt Cave along the base line. 



POST-PLEISTOCENE MAMMALS 241 

flaviventris, Neotoma cinerea, Sorex cinereus, Lagurus curtatus, and 
Microtus ochrogaster that are now found farther north and east and at gen- 
erally higher altitudes or in more humid areas (Stearns 1942; Schultz and 
Howard 1935; Murray 1957; Harris 1970). Very little is known about the 
faunal transition from Late Wisconsin to the present, although Murray 
(1957) outlines a post-Wisconsin warming and drying trend. Studies of 
amphibian and reptile remains from Pratt Cave (Gehlbach and Holman 
1974) and of the pollen (Schoenwetter, unpublished; Bryant, unpublished) 
show a minor change toward more arid conditions through the time 
represented by these deposits. 

METHODS 

The excavations were done according to standard archeological tech- 
niques. A base line (compass orientation N41°E, S41°W) was established 
along the long axis of the cave 36 in. (90 cm) above the highest point of the 
cave deposits. A grid of 3-ft 2 (234 cm 2 ) blocks was used for the excava- 
tions, and material was removed in 6-in. (15 cm) levels. Level data within 
the deposits were recorded with reference to the base line. 

Mammal bones were identified to the lowest possible taxon. For most 
mandibles and maxillae with teeth, specific identifications were possible. 
Most post-cranial material could not be identified below the generic level. 
Environmental interpretations are based on modern distributions and 
habitat preferences of the species represented in the cave deposits. 

The mammalian remains are cataloged in the collection of the Texas 
Memorial Museum, the University of Texas at Austin, under the locality 
number TMM 41172. Abbreviations used in the text and tables are as 
follows: TMM, Texas Memorial Museum; M, Texas Memorial Museum 
Recent Collection; UNM, University of New Mexico; TNHC, Texas 
Natural History Collection (housed at the Texas Memorial Museum). 

PRESERVATION OF BONES 

Mammal bones from Pratt Cave are, in general, well preserved. Except for 
bones of large animals, there is little breakage. There is a small amount of 
articulated material, some of which (e.g., the foot of a small mammal) is held 
together by dried ligaments. Most bones are of animals no larger than a jack- 
rabbit, and the low degree of breakage in this size range suggests an owl 
pellet accumulation, although some species (e.g., wood rats) certainly lived 
in and around the cave. 

There are some burned human bones which are the remains of an 
aboriginal cremation. The bones of the larger animals such as deer, antelope, 
and horse were probably brought into the cave by mammalian carnivores, 
for the scarcity of artifacts in the cave rules out any significant occupation by 
humans. 



242 LUNDELIUS 

AGE 

The age of the deposits in Pratt Cave is only partially known. Four C-14 
dates are available: TX-1021 wood from a cache in the back of the cave 1420 
±60 years BP; TX-1022 basketry from the cache at the back of the cave 1840 
±60 years BP; TX-2191 bone from level 2 of blocks 10-15 apatite carbonate 
2560 ± 340 years BP, collagen 2820 ± 180 years BP; TX-2 192 bone from level 
2 of blocks 2-9 apatite carbonate 2090 ± 420 years BP, collagen 2320 + 70 
years BP. 

Both dates based on the basketry and wood from the cache are younger 
than those based on bone from level 2. This is consistent with the higher 
stratigraphic position of the cache. 

The C-14 dates from the two areas of level 2 differ by approximately 500 
years. The dates based on the apatite carbonate fraction overlap extensively 
at 1 standard deviation. Those based on the collagen fraction do not overlap 
at 2 standard deviations. It is doubtful that a significant age difference exists 
for level 2 in blocks 2 through 15. 

These dates are all from the back part of the cave where the deposits are 
relatively thin. The correlation of these dated units with the sequence near 
the opening is tentative. A lithologic correlation indicates that the tan silty 
layer found in level 2 in blocks 5-7, 1 3- 1 5 is equivalent to a similar silt found 
in level 6 in blocks 22-24. In all these areas this silt rests on the bedrock and 
appears to represent the beginning of the accumulation of sediment in the 
cave. 

ENVIRONMENTAL INTERPRETATION 

The Pratt Cave mammalian fauna has no extinct taxa. Three species, 
Marmota flaviventris, Neotoma cinerea, and Geomys bursarius, are not 
known to live today in the southern Guadalupe Mountains. The first two 
species are potential indicators of climatic change. 

Both Marmota flaviventris and Neotoma cinerea now live well to the 
north of Pratt Cave and at higher altitudes. Harris (1963a, 1963b, 1970) 
showed that the limit of the southern and lower altitudinal distribution of 
Marmota flaviventris is controlled by the availability of green fodder during 
the spring, which is, in turn, controlled by winter precipitation. 

The factors that control the distribution of N. cinerea are less well known 
and more complex than in the case of Marmota because they apparently 
involve competing species of Neotoma. The situation is further confused by 
the fact that the N. cinerea material from Pratt Cave (and also from Dark 
Canyon Cave — a late Pleistocene assemblage near Carlsbad, New Mexico) 
is closer in size to living populations in Wyoming than to living populations 
in New Mexico (Table 1). Moreover, there are evidently four species of 
Neotoma in the Pratt Cave Fauna. Finley ( 1958) found no area in Colorado 
where the same four species occurred together and only small areas where 
three were found together. The lack of stratification in Pratt Cave and the 
presence of an intrusive burial make demonstration of contemporaneity of 



POST-PLEISTOCENE MAMMALS 



243 



TABLE 1. Numerical data on occlusal length of Mi in various samples of Neotoma. 



Species and locality 



No. of ., Observed 

Mean 
specimens range 



N. cinerea (Pratt Cave) 

N. cinerea (Recent-New Mexico) 

N. cinerea (Recent- Wyoming) 

N. cinerea (Pleistocene- Dark Canyon Cave) 

N. mexicana (Recent- New Mexico) 



3 


3.72 


3.63-3.81 


4 


3.24 


3.08-3.43 





3.75 


3.52-3.96 


4 


3.63 


3.38-3.86 


2 


3.20 


3.16-3.24 



all four species difficult if not impossible. N cinerea specimens were 
recovered from blocks 13-15, level 1, along with N. albigula and N. 
mexicana, from blocks 10-12, level 2, with TV. albigula, N. micropus, and N. 
mexicana; and from level 1 in the hearth with N. mexicana and N. albigula. 
The material in part of blocks 13-15 had been disturbed, and the association 
of the remains of the three species might be the result of this disturbance. 
However, the material in blocks 10-12 showed no evidence of disturbance, 
and the association may be more meaningful. In addition, three species of 
Neotoma (N. mexicana, N albigula, and N micropus) are found together in 
many blocks at many levels, and it is unlikely that this represents post- 
depositional mixing. 

The presence of Marmota flaviventris and Neotoma cinerea in the Pratt 
Cave deposits indicates more mesic conditions at some time in the past. The 
four species of Neotoma indicate a greater diversity of niches in the past. 
This is similar to the situation found in Pleistocene faunas in many areas of 
North America (Hibbard 1960; Guilday et al. 1964; Lundelius 1967; 
Dalquest et al. 1969). In the absence of a more complete post-Pleistocene 
sequence, it is difficult to ascertain whether this diversity is a Pleistocene 
relict or whether it represents a short-lived situation in the transition from 
Pleistocene to present conditions. The distribution of Neotoma in the Pratt 
Cave deposits contributes little information on this question because of the 
absence of well-defined stratigraphy and sufficient C-14 dates to permit 
reliable correlations. The specimens of N. cinerea are either associated with 
the C-14 date of 2560 years BP or are stratigraphically above it. The failure 
to find all four species of Neotoma together in any late Pleistocene fauna so 
far reported elsewhere in the southwest might be taken as evidence of such a 
transition. A careful restudy of the Neotoma material from all faunas in this 
region is needed. 

McKittrick Canyon today is much more mesic than the surrounding area. 
It seems likely that this has been the case for all of post-Pleistocene time and 
perhaps earlier. Conditions in the canyon remained favorable for the 
maintenance of both Marmota flaviventris and Neotoma cinerea after they 
had disappeared in the surrounding area. Eventually, this regional climatic 
change affected McKittrick Canyon fauna. 



244 LUNDELIUS 

Geomys bursarius is no longer found in the region of the Guadalupe 
Mountains or in the area to the south. This is the area that separates G. 
bursarius from its close relative G. arenarius in the El Paso region. The 
presence of G. bursarius in level 1 of blocks 10-12 of the Pratt Cave deposits 
indicates a recent disappearance from this locality (post 2000 years). The 
reason for its disappearance is not known. Semken (1961) postulated soil 
erosion resulting from overgrazing during the last century as the cause of this 
animal's recent disappearance from the Edward's Plateau of central Texas. 

SYSTEMATIC DESCRIPTIONS 

CLASS Mammalia 

Order Insectivora 
Family Soricidae 

Notiosorex crawfordi (Coues) 
Order Carnivora 
Family Mustelidae 

Mephitis mephitis (Schreber) 

Mustela frenata Lichtenstein 
Family Canidae 

Urocyon cinereoargenteus (Schreber) 
Family Procyonidae 

Bassariscus astutus (Lichtenstein) 

Procyon lot or (Linnaeus) 
Order Perissodactyla 
Family Equidae 

Equus sp. 
Order Artiodactyla 
Family Cervidae 

Odocoileus sp. 
Family Antilocapridae 

gen. and sp. indet. 
Order Rodentia 
Family Sciuridae 

Spermophilus variegatus (Erxleben) 

Cynomys ludovicianus (Ord) 

Eutamias canipes Bailey 

Marmota flaviventris (Audubon and Bachman) 

gen. and sp. indet. 
Family Erethizontidae 

Erethizon dorsatum (Linnaeus) 
Family Cricetidae 

Microtus mexicanus (Saussure) 

Neotoma albigula Hartley 

N. cinerea (Ord) 

N. mexicana Baird 

N. micropus Baird 

Onychomys leucogaster (Wied-Neuwied) 

O. torridus (Coues) 

Peromyscus difficilis (Allen) 

Peromyscus sp. 



POST-PLEISTOCENE MAMMALS 245 



Reithrodontomys megalotis (Baird) 

Sigmodon hispidus Say and Ord 
Family Geomyidae 

Pappogeomys castanops (Baird) 

Geomys bursarius (Shaw) 

Thomomys bottae (Eydoux and Gervais) 
Family Heteromyidae 

Dipodomys ordii Woodhouse 

D. spectabilis Merriam 

Perognathus hispidus Baird 

P. merriami Allen 
Order Lagomorpha 
Family Leporidae 

Lepus californicus Gray 

Sylvilagus audubonii (Baird) 

Sylvilagus sp. 



CLASS Mammalia 

ORDER Insectivora 
Family Soricidae 
Notiosorex crawfordi (Coues), Desert Shrew 

Material.— Nine right and three left mandibles (TMM 41 172-63, -373, -400, -659, -762, -763, 
-765, -766, -768 and TMM 41 172-42, -761, -767); two palates (TMM 41 172-759, -764); one 
right maxillary (TMM 41172-760) from levels 1 through 5. 

Description and Discussion. — This material is readily recognized as N. crawfordi by the 
presence of the deep, rounded notch in the condyle of the mandible (Hibbard and Taylor 1960) 
and the presence of only three unicuspid teeth in the upper dentition. In addition, the teeth are 
only lightly pigmented, and the coronoid process lacks a posterointernal ramal fossa (Gaughran 
1954). 

The desert shrew is known from widely scattered localities in the southwestern desert, grass- 
land, and woodland communities. 

ORDER Carnivora 
Family Mustelidae 

Mephitis mephitis (Schreber), Striped Skunk 

Material.— One left tibia (TMM 41172-510) from blocks 10-12, level 1; one left mandible 

(TMM 4 1 1 72-222) from blocks 10-12, level 2; left M , (TM M 4 1 1 72-773) from blocks 2-4, level 

2. 

Description and Discussion. — This material shows no differences from the corresponding parts 

of Recent specimens. The striped skunk is present in the area surrounding Pratt Cave but is less 

common there than the hognose skunk. 

Mustela frenata Lichtenstein, Long-tailed Weasel 

Material.— One left maxillary with P 4 -M l and alveolus for P 3 (TMM 41 172-1 153) from block 
10-12, level 2; one left Mi (TMM 41172-327) from level 1 of the hearth. 
Description and Discussion. — A comparison of the Pratt Cave material with a Recent specimen 
from Ohio (TMM M-1751) shows only minor differences. The protocone of the P 4 and the 
entire M 1 of the Pratt Cave specimen are slightly larger than in the Ohio specimen. In addition, 
the metacone and the posterior part of the stylar area of the M 1 are larger in the Pratt Cave 
specimen. These differences are minor and are no more than expected between specimens from 



246 LUNDELIUS 

such widely separated localities. The M , cannot be distinguished from that of the Recent 
specimen. 

The long-tailed weasel is widely distributed in the United States and Mexico, except for 
southwest Arizona, southeast California, and Baja California, and is recorded from Culberson 
County (Davis 1960: 80). 

Family Canidae 

Urocyon cinereoargenteus (Schreber), Gray Fox 

Material.— One left M 2 (TMM 41 172-326) from level 1 of the hearth. 
Description and Discussion.— The Pratt Cave specimen is considerably smaller than the 
corresponding tooth of Vulpes vulpes and is somewhat smaller than the M 2 of U. 
cinereoargenteus fromcentral Texas.The tooth resembles the corresponding tooth of Recent U. 
cinereoargenteus in the strong development of the outer cingulum, the relatively low paracone 
and metacone, and the broad inner area. In addition, the angle formed by the intersection of a 
line joining the paracone and metacone with one joining the protocone and metaconule is much 
greater in U. cinereoargenteus than in V. macrotis. The length along the ectoloph is 7.5 mm; the 
width normal to the ectoloph is 8.6 mm. 
This species is common around Pratt Cave today. 

Family Procyonidae 

Bassariscus astutus (Lichtenstein), Ringtail 

Material.— One left P 4 (TMM 41172-372) from blocks 13-15, level 2; one M 2 (TMM 
41 172-771) from blocks 8-9, level 1; one canine (TMM 41 172-776) from blocks 8-9, level 2; one 
left *iP 4 (TMM 41172-777) from blocks 25-27 B, level 4. 

Description and Discussion. — These specimens cannot be distinguished from the homologous 
teeth of the Recent form. This is a widespread upland species in the southwestern United States 
and is common in the Guadalupe Mountains today. 

Procyon lotor (Linnaeus), Racoon 

Material.— Anterolabial part of one right P 4 (TMM 41172-649) from blocks 10-12, level 1. 
Description and Discussion. — The P 4 has the parastyle, the anterior half of the paracone, and 
about one-fourth of the protocone. Although the shape and size of the parastyle of P 4 is variable 
in this species, the Pratt Cave specimen is almost identical to a Recent specimen from Sutton 
County, Texas (TMM M-2340). 
The racoon is common along McKittrick Canyon today. 

ORDER Perissodactyla 
Family Equidae 
Equus sp. 

Material.— The distal end of one metapodial (TMM 41 172-677) from level 1, blocks 16-18 R 
and the distal articular surface of one phalanx (TMM 41 172-554) from level 3, blocks 26-28 B. 
Description and Discussion.— These two specimens are the only indication of horses recovered 
from the Pratt Cave deposits to date. They represent a small form about the size of an ass. The 
metapodial has been slightly burned and has the anterior (dorsal) surface of the shaft broken 
away. It is not possible to determine on the basis of morphology whether these specimens were 
derived from a modern feral or domestic horse or from an extinct Pleistocene horse, but their 
positions in the cave and their preservation indicate they represent the former. 

Both specimens were recovered from the topmost unit of the cave. In view of the C- 14 dates of 
2090 and 2560 years BP from level 2 of blocks 2 to 1 5, they apparently represent modern feral or 
domesticated horses. 



POST-PLEISTOCENE MAMMALS 247 

ORDER Artiodactyla 
Family Cervidae 
Odocoileus sp. 

Material.— Two incisors and fragments of postcranial skeleton from blocks 5-7, level 2; blocks 
8-9, level 2 and silt layer of level 2; and blocks 25-27 C, level 4. 

Description and Discussion. — This genus, as with most of the larger animals, is not well 
represented in the deposits of Pratt Cave. Both O. virginianus and O. hemionus are found in this 
region today, but the latter is much more common. 

ORDER Rodentia 
Family Sciuridae 
Spermophilus variegatus (Erxleben), Rock Squirrel 

Material. — Maxillaries, mandibles, and molars from blocks 10-12, level 2; blocks 19-21 B, level 
2; blocks 22-24 A, level 2; blocks 22-24 B, levels 2, 5, 6; blocks 25-27 B, levels 5, 6; blocks 25-27 
C, levels 3, 4. 

Description and Discussion.— The material from Pratt Cave cannot be distinguished from 
Recent specimens in size or morphology. The limestone cliffs in McKittrick Canyon are typical 
of their habitat, and the rock squirrel is common there today. 

Cynomys ludovicianus (Ord), Black-tailed Prairie Dog 

Material.— Posterior part of one left mandible (TMM 41 172-217) from level 3 of blocks 25-27 
C; left Mi (TMM 41 172-1324) from level 5 of blocks 23-25 B; lower molar (TMM 41 172-751) 
level 2 of blocks 8-9; lower molar (TMM 41 172-755) from level 2 of blocks 4-5. 
Description and Discussion. — The mandible is indistinguishable from the same part of the 
mandible of Recent specimens. 

The historic range of this species included the area of Pratt Cave, although the nearest 
extant colony is some 20 miles distant. Man has been the most important agent in Recent range 
restrictions of the black-tailed prairie dog. 

Eutamias canipes Bailey, Gray-collared Chipmunk 

Material. — One right mandible with Mi (TMM 41172-753) from level 2 of blocks 2-4. 
Description and Discussion. — The Pratt Cave specimen resembles specimens of E. canipes in 
the Texas Natural History Collection (TNHC 135, 137) from New Mexico in size and 
morphology of the M i . The metaconid is strongly joined to the anterior cingulum which is well 
defined and straight. The protoconid and hypoconid are well separated, and a well-developed 
mesoconid is located between them. 

The specimen differs from Spermophilus spilosoma in its smaller size, the presence of a well- 
defined mesoconid, and in the subdivisions of the trigonid basin by a ridge that extends from the 
protoconid to the anterior cingulum. It differs from Ammo spermophilus interpres in its slightly 
smaller size, presence of a well-defined mesoconid, and by the divided trigonid basin. In addi- 
tion, the metalophid is better developed in E. canipes, and there is no well-developed cuspule on 
the anterior cingulum. The metaconid of A. interpres is more isolated from the protoconid and 
the cingular cuspule than is the case in E. canipes. 

The gray-collared chipmunk is found in the higher parts of the Guadalupe Mountains, 
including The Bowl and Dog Canyon (Hall and Kelson 1959). It is doubtful that it signifies any 
change in conditions in McKittrick Canyon. 

Marmota flaviventris (Audubon and Bachman), Yellow-bellied Marmot 

Material.— Labial part of one left M 3 (TMM 41 172-742) from level 2, blocks 5-7, associated 
with C-14 date TX-2192. 

Description and Discussion.— The Pratt Cave specimen differs only in minor details from the 
M 3 of a Recent specimen from Colorado (TMM M-1248). The metaloph is slightly more 



248 LUNDELIUS 

complicated in the Pratt Cave specimen, with a slightly better-developed metaconule. The 
length along the ectoloph of the Pratt Cave specimen is 5.6 mm; that of the Colorado specimen 
is 5.2 mm. 

The occurrence of the yellow-bellied marmot in the Pratt Cave deposits indicates that some 
change has taken place in the environment of McKittrick Canyon since it lived there. The 
marmot today is found in the high mountain forests of New Mexico. Its presence in the 
McKittrick Canyon deposits could imply the existence of similar forests in the southern Guada- 
lupe Mountains at some time in the past. However, Harris and Findley (1964) have pointed out 
that this animal lives in other habitats and have interpreted its presence in a late Pleistocene 
fauna from north -central New Mexico as indicating open conditions similar to those found 
today in central Wyoming. 

Harris (1970) postulated, on the basis of its distribution in New Mexico, that the primary 
limiting factor in the southward and lower altitude distribution of the marmot is the avail- 
ability of sufficient precipitation during the winter to provide abundant green fodder in the 
spring prior to the summer rains. The mountains in the southern part of New Mexico and 
Arizona in which marmots would be expected, but are absent, are separated from the northern 
mountain masses by dry areas. Harris believes this would block repopulation. He further 
postulated that the elimination of marmots from these mountains took place after the inter- 
vening areas had become dry, or repopulation would have taken place. This interpretation 
(Harris 1970; Harris and Findley 1964) has implications for the age of the Pratt Cave deposits as 
well as the environmental conditions. The earlier interpretations of the climatic significance of 
marmots in fossil and subfossil assemblages made by Stearns (1942) and Murray (1957) imply 
that most of the southward extensions of range of the marmot in the past date from the late 
Pleistocene. The interpretation of Harris that marmots survived in isolated favorable areas for 
significant lengths of time after the end of the Pleistocene is confirmed by the Pratt Cave 
specimen which is associated with a C-14 date of 2090 ± 420 years BP (TX-2192). The upper 
parts of McKittrick Canyon are still quite mesic and the lower portion may have been so, too, 
until quite recently. 

Family Sciuridae — genus and species undetermined 

Material.— One edentulous right maxillary (TMM 41172-144) from level 4 of blocks 25-27. 
Description and Discussion. — This specimen is from a small sciurid about the size of 
Ammospermophilus interpres but is not generically or specifically identifiable. 

Family Erethizontidae 

Erethizon dorsatum (Linnaeus), Porcupine 

Material.— Two quills (TM 41172-1106) from block 5-7, level 2. 

Description and Discussion.— This is the only material of this species recovered from the Pratt 

Cave deposits. This animal is widespread in west Texas and New Mexico today where it 

frequently uses crevices and caves as shelters. It is part of the Recent fauna of the Guadalupe 

Mountains. 

Family Cricetidae 

Microtus mexicanus (Saussure), Mexican Vole 

Material.— Two left M^s (TMM 41 172-397, -398) irom level 2, blocks 19-21. Two left man- 
dibles (TMM 41172-1332, 1333) and a left maxillary (TMM 41172-1334) from level 2 of 
blocks 22-24 A. 

Description and Discussion. — These two teeth are the only remains of microtines found to date. 
One tooth (TMM 41 172-397) has three closed triangles and a fourth (inner) triangle which is 
confluent with the posterior loop. The other specimen (TMM 41 172-398) has only two closed 
triangles and a third (outer) which is confluent with the posterior loop. Both conditions are 
found in M. mexicanus from Otero County, New Mexico (TNHC 2284, 2287, 2289). The M 3 of 
M. pennsylvanicus also shows three closed triangles and a fourth confluent with the posterior 



POST-PLEISTOCENE MAMMALS 249 

loop. All 12 specimens of M. pennsylvanicus examined showed the three closed triangles on M 3 . 
The teeth from Pratt Cave are referred to M. mexicanus. 

In the southern Guadalupe Mountains this species is now restricted to elevations above 7000 
ft, where it is found in grassy areas in or near coniferous forest. It is possible that in the recent 
past it was found at lower elevations in McKittrick Canyon and that the present restriction is a 
recent change, similar to the recent disappearance of Pitymys pinetorum from many areas of 
central Texas in the last 1000 years (Lundelius 1967). It is also possible that the remains of this 
species in Pratt Cave were brought from higher elevations by owls. Their rarity would support 
this hypothesis. 

Neotoma albigula Hartley, White-throated Woodrat 

Material. — Mandibles and Mi 's from blocks 2-4, level 2; blocks 8-9, level 2; blocks 10-12, 
levels 1, 2; blocks 13-15, level 1; blocks 19-21 A, levels 2, 3; blocks 19-21 B, levels 1,3; blocks 
22-24 A, levels 2, 3; blocks 22-24 B, level 5; blocks 22-24 C, level 5; blocks 25-27 A, level 2; 
blocks 25-27 B, levels 4, 5; blocks 25-27 C, levels 3, 4, 5; level 1 of the hearth. 
Description and Discussion. — This material is assigned to this species on the basis of the 
absence of a dentine tract on the anteroexternal face of M , and because the width of the sec- 
ond loph of M, is less than 1.94 mm. The width of the second loph on the M, is not an 
absolutely certain criterion upon which to base the differentiation of N. albigula and N. 
micropus; hence, the possibility exists that not all specimens are correctly assigned. 

This species is widely sympatric with N. micropus, including the Guadalupe region. Finley 
(1958) found that in Colorado, where these two species are sympatric, N. albigula is restricted to 
rocky uplands and N. micropus, to more open lowlands. This same situation was reported by 
Bailey (1905) for west Texas and holds in the Guadalupe Mountains. N. albigula is rare in 
McKittrick Canyon, however, and N. micropus has not been trapped there. 

Neotoma cinerea (Ord), Bushy-tailed Woodrat 

Material.— One right mandible with M1-2 (TMM 41 172-229) from level 2 of blocks 10-12; one 
right M, (TMM 41172-1104), one right M 1 (TMM 41172-1102), one right M 2 (TMM 
41 172-1 105) from levels 1 of blocks 13-15; one left and one right mandible (TMM 41172-283, 
281) from level 1 of the hearth. 

Description and Discussion. — The specimens listed above have dentine tracts on the molars 
which are as well developed as in Neotoma mexicana but are substantially larger. The hori- 
zontal ramus of the mandible (TMM 41 172-229) is deeper and more massive than that of N. 
mexicana. The basal closure of the labial re-entrants of the M] , M',and M 2 is better developed 
than in N. mexicana or N. micropus. The M j 's from Pratt Cave are not only larger than those of 
N. mexicana from New Mexico but are also larger than four modern specimens of N. cinerea 
from New Mexico (UNM 27795, 17860, 16250, 94). They are about the same size as the Mi's 
from a sample of N. cinerea from Wyoming and specimens of N. cinerea from a late Pleistocene 
deposit from Dark Canyon Cave near Carlsbad, New Mexico (Table 1). 

The difference in size between the modern N cinerea from New Mexico and Wyoming 
suggests a north-south size cline in this species today. The large size of the Dark Canyon Cave 
specimens indicates that during the Pleistocene, large-sized N. cinerea were living farther south 
than today. The Pratt Cave specimens show that at least one population retained large size well 
into the Holocene. 

The presence of this taxon, like Marmota, is unexpected in a fauna this young. However, it 
has been reported in late Pleistocene faunas from Williams Cave (Ayer 1937), Dry Cave (Harris 
1970), and Burnet Cave (Murray 1957). 

Neotoma mexicana Baird, Mexican Woodrat 

Material. — Numerous mandibles and Mfs from blocks 2-4, level 2; blocks 5-7, level 2; blocks 
8-9, level 2; blocks 10-12, levels 1, 2; blocks 13-15, levels 1, 2; blocks 16-18 A, level !; blocks 
16-18 B, level 1; blocks 19-21 A, levels 1, 2, 3; blocks 19-21 B, levels 1, 2, 3; blocks 22-24 A, 



250 LUNDELIUS 



levels 1, 2, 3, 4; blocks 22-24 B, levels 2, 4, 5, 6; blocks 22-24 C, levels 2, 4; blocks 25-27 A, level 
2; blocks 25-27 B, levels 3, 4, 5, 6; blocks 25-27 C, levels 3, 4, 5, 6; blocks 25-27 D, levels 3, 4; 
level 1 of the hearth, blocks 20-21 B, level 4, and the packrat nest of block 19. 
Description and Discussion. — Three species of Neotoma (N. mexicana, N. albigula, and N. 
micropus) have been reported from the Guadalupe Mountains. In addition, N. cinerea, which 
now occurs in northern New Mexico, and N.floridana, which now occurs in the eastern half of 
Texas, might be expected to have extended their ranges southward and westward, respectively, 
during the last glacial stage. 

The specific identification of Neotoma mandibles and teeth is difficult. Recently, Dalquest et 
al. (1969) have suggested several criteria by which N.floridana, N. micropus, and N. albigula 
can be separated. According to them, the greater size of N.floridana is reflected in a greater 
breadth of the molar rows (greater than 8.7 mm) than either N. micropus or N. albigula. This 
character was checked in 32 Recent specimens of N. albigula from the Texas High Plains and 30 
Recent specimens of N. micropus from south Texas in the Texas Natural History Collection. 
None of these specimens had a molar row breadth exceeding 8.6 mm. None of the few palates 
present in the Pratt Ca\e collection has a molar row breadth equal to 8.7 mm. As a result, there 
is no evidence of the presence of N floridana from Pratt Cave. 

Dalquest et al. (1969) separated N. micropus from N. albigula on the basis of the width of the 
second lophid of M l . They found that the width of this lophid in N. micropus always exceeded 
1 .94 mm, whereas in N albigula it was always less than 1 .94 mm. However, this was checked in 
the Texas Natural History Collection specimens of N. albigula and N. micropus mentioned 
above with the following results: in five specimens of N. micropus, the width of the second 
lophid of M , is less than 1 .94 mm; in three specimens of N. albigula, this width is greater than 
1.94 mm. Thus this character is not completely reliable for differentiating N. micropus and N. 
albigula. 

Neotoma mexicana was identified on the basis of the dentine tract on the anteroexternal side 
of the M i . The dentine tract on Mj extends one-fourth to one-third the distance from the root 
to the crown of an unworn tooth. The dentine tract on the M 2 is shorter. An examination of 27 
Recent specimens of N. albigula from Dawson County, Texas, in the Texas Natural History 
Collection shows a short, incipient dentine tract in two specimens, TNHC 3 144 and 1038. This 
could not be confused with the situation in either N. cinerea or N. mexicana. None of the Pratt 
Cave specimens shows the enamel islands which result from isolation of the inner parts of the re- 
entrant folds in deeply worn teeth as is seen in N. cinerea. 

N. mexicana is common in McKittrick Canyon today where it occupies cracks and crevices in 
the canyon walls. 

Neotoma micropus Baird, Gray Woodrat 

Material.— Mandibles and Mi's from blocks 2-4, level 2; blocks 5-7, level 2, blocks 8-9, levels 
1,2; blocks 10-12, level 2; blocks 13-15, level 2; blocks 19-21 A, level 1 ; blocks 22-24 A, level 2; 
blocks 22-24 C, level 5; blocks 25-27 B, levels 4, 5; blocks 25-27 C, level 5. 
Description and Discussion.— This material is assigned to this species on the basis of the 
absence of a dentine tract on the anteroexternal face of M i and a width of the second loph of M i 
more than 1.94 mm. The general ecological preferences and relationships to other species have 
been discussed above. 

Onychomys leucogaster (Wied-Neuwied), Northern Grasshopper Mouse 

Material.— One left mandible with Mj (TMM 41 172-1304) from level 2, blocks 8-9; one eden- 
tulous right mandible (TMM 41172-535) from level 6, blocks 26-28. 

Description and Discussion. — These specimens have the prominent, backwardly curved 
coronoid process of Onychomys, but their M j_ 3 alveolar lengths are larger than those of either 
of the other Pratt Cave specimens or the Recent sample of O. torridus (Tables 2-3). The large 
size of these specimens suggests that they are probably O. leucogaster, and they are tentatively 
assigned to this species on that basis. 



POST-PLEISTOCENE MAMMALS 



25: 



TABLE 2. Numerical data on Onychomys from Pratt Cave. 



Specimen number 


LMi 


L lower alveolus 

Mt.3 






Onychomys torridus 


TMM 41172-411 


1.65 


3.85 


TMM 41172-99 


- 


3.76 


TMM 41172-100 


- 


3.72 


TMM 41172-193 




Onychomys leucogaster 


TMM 41172-535-1304 


- 


4.58 




1.97 


4.03 



LM 1 



L upper alveolus 
M l-3 



1.93 



4.09 



TABLE 3. Numerical data of Recent samples of Onychomys leucogaster and Onychomys 
torridus, San Patricio County, Texas. 



Sample 
size 



Observed 
range 



Mean ± 

standard 

error 



Standard 
deviation 



Coefficient 

of 
variation 



Onychomys leucogaster 



LMi 


9 


1.72-2.01 


1.88+0.032 


0.096 


5.11 


Alveolus LM |_3 


12 


3.77-4.48 


4.19±0.065 


0.225 


5.37 


LM 1 


8 


2.06-2.23 


2.15±0.023 


0.065 


3.01 


Alveolus LM 1_3 


13 


4.35-4.89 


4.71 + 0.044 


0.159 


3.38 






Onychomys torridus 






LMi 


2 


1.56-1.67 


1.62 






Alveolus LM i_ 3 


2 


3.78-3.79 


3.785 






LM 1 


2 


1.94-2.01 


1.98 






Alveolus LM 1_3 


2 


3.96 


3.96 







Onychomys leucogaster is known to the north of Pratt Cave in the vicinity of Carlsbad, New 
Mexico (Hall and Kelson 1959: 663). It is likely that its distribution extended farther to the 
southwest during the Pleistocene, and McKittrick Canyon provided a refuge into Holocene 
time. 

Onychomys torridus (Coues), Southern Grasshopper Mouse 

Material.— One right maxillary with M 1 (TMM 41172-193), one right mandible with Mj 
(TMM 41172-411), one edentulous right mandible (TMM 41172-100), one edentulous left 
mandible (TMM 41172-99), all from level 2, blocks 19-21. 

Description and Discussion. — All of the mandibles have the coronoid process strongly recurved 
posteriorly. The Mi and M 1 are simple, with the high, widely exposed principal cusps charac- 
teristic of Onychomys. 

This material is assigned to O. torridus on the basis of size. The lengths of Mi , M 1 , and of 
alveoli for M1-3 and M 1 " 3 are within the range of size of a Recent sample of O. torridus from 



252 LUNDELIUS 

Trans-Pecos Texas. Only one of these measures, the M1-3 alveolar length, overlaps the lower 
end of the size range of a sample of O. leucogaster from south Texas (Tables 2-3). 
This species is present in the general area of Pratt Cave today. 

Peromyscus difficilis (Allen), Zacatecan Deer Mouse 

Material. — One right mandible with M1-2 from the hearth area (TMM 41 172-299), one right 
and one left mandible with M1-2 from level 1 (TMM 41 172-585, -1201); one right and two left 
mandibles with M1-2 from level 2 (TMM 41172-1157, -1206, -97); one right mandible with 
Mi_ 2 from level 3 (TMM 41 172-1266). 

Description and Discussion. — All Peromyscus mandibles with low incisor root capsules are 
heterogeneous as to size and complexity of the teeth. The specimens listed above are large in size 
(length of M i_2 greater than 3.26 mm) and have stylids and lophids present on both Mi and M 2 . 
The length of M1-2 of a Recent sample of P. difficilis (N = 5) ranges from 3.16 mm to 3.43 mm, 
and both teeth are complicated with lophids and stylids as reported by Martin (1968). 

Peromyscus sp. 

Material. — Numerous mandibles from levels 1, 2, 3, 4, and 6. 

Description and Discussion. — The Peromyscus material has presented problems of identifica- 
tion which have not been solved to the author's satisfaction. The extensive overlap in size and 
dental characters of a number of species makes definite identification of individual specimens 
very difficult. 

Dalquest et al. (1969) have used the size of the capsule housing the root of the lower incisor to 
separate P. leucopus and P. maniculatus from P. boylii and P. pectoralis. They found that this 
capsule formed a prominent bulge on the labial side of the mandible in the first pair of species 
and little or no bulge in the second. In addition, an examination of mandibles of P. difficilis and 
P. truei shows that the former also has an inconspicuous capsule and that the latter has a 
slight to moderately bulged capsule. 

The mandibles from Pratt Cave can be easily divided into two groups on the basis of the size 
of this capsule. However, it is clear from an examination of scatter diagrams of length versus 
width of the lower molars and frequency histograms of the length of Mi_2 that both groups are 
heterogeneous. 

Dalquest et al. (1969) pointed out that in central and east Texas P. maniculatus can be dis- 
tinguished from P. leucopus by its shorter Mi_2 but that this character does not hold for the 
populations in west Texas. Indeed, I have been unable to find reliable characters to differen- 
tiate these populations in west Texas. 

The distribution maps of Hall and Kelson (1959) showed both species occurring in the 
Guadalupe Mountains, but Davis (1940) and Davis and Robertson (1944) reported only the 
occurrence of P. leucopus. The inability to separate these species makes it impossible to 
determine whether P. maniculatus was present in the Guadalupe Mountains in the past. 

Those mandibles with small incisor root capsules and teeth with a low degree of development 
of stylids and lophs show some variation and may be heterogeneous as to species. Many are 
similar in size and morphology to P. boylii rowleyi now found in west Texas and probably 
represent that species. 

Reithrodontomys megalotis (Baird), Western Harvest Mouse 

Material.— Mandibles and maxillaries from blocks 5-7, level 2; blocks 8-9, levels 1,2; blocks 
10-12, levels 1, 2; blocks 16-18, level 1; blocks 19-21 A, level 2; blocks 19-21 B, level 3; blocks 
22-24 A, level 3; blocks 22-24 C, levels 4, 5; packrat nest in block 19. 
Description and Discussion. — The first primary fold of M 3 is shorter than the second primary 
fold and does not extend half way across the tooth. The labial folds on the M 12 are tight and V- 
shaped when viewed from the side. 

R. megalotis is present in the Bowl area, nearly 1000 ft above McKittrick Canyon today. It is 
similar morphologically to R. montanus, which is also recorded from Trans-Pecos Texas 



POST-PLEISTOCENE MAMMALS 253 

(Hooper 1952; Hall and Kelson 1959). All specimens of Reithrodontomys from Pratt Cave 
more closely resemble R. megalotis than R. montanus. 

Sigmodon hispidus Say and Ord, Cotton Rat 

Material.— Numerous mandibles and maxillaries from blocks 2-4, level 2; blocks 8-9, level 2; 
blocks 10-12, levels 1, 2; blocks 19-21 A, level 2; blocks 19-21 B, level 3; blocks 22-24 A, levels 
2, 3; blocks 22-24 C, levels 3, 4; blocks 25-27 B, level 5; blocks 25-27 C, level 4; level 1 of the 
hearth. 

Description and Discussion. — All of the Sigmodon material is referable to S. hispidus. A 
comparison of the Pratt Cave material with S. ochrognathus reveals that it is too large to be 
readily assigned to that species. It falls within the size range of a sample of S. hispidus from 
central Texas. 

The cotton rat is widely distributed in the southern part of the United States. It is reported by 
the Guadalupe Mountains Park Staff (1965), and its presence in the Pratt Cave deposits is not 
unexpected. 

Family Geomyidae 

Pappogeomys castanops (Baird), Yellow-faced Pocket Gopher 

Material.— Mandibles and maxillaries from blocks 2-4, level 2; blocks 8-9, levels 1 and silt layer 
of 2; blocks 10-12, level 1 ; blocks 19-2 1 A, level 3; blocks 22-24 A, levels 1,2,3; blocks 22-24 B, 
levels 2, 4; blocks 25-27 B, levels 3, 4, 5, 6; blocks 25-27 C, level 4; the hearth. 
Description and Discussion. — The material referred to this species cannot be distinguished 
from Recent specimens. The size is large, the upper incisors are unisulcate, and M|_3 have 
enamel plates only on the posterior side. 

Pappogeomys castanops in the Pratt Cave deposits is expected. This gopher occupies deep 
soils relatively free of rocks (Davis 1960). In the vicinity of Pratt Cave this habitat is repre- 
sented by the terrace deposits that flank the existing creek, and P. castanops is usually present 
there today. However, the periodic floods that sweep McKittrick Canyon often decimate the 
population. 

Geomys bursarius (Shaw), Plains Pocket Gopher 

Material. — One right mandible (TMM 41 172-26) from level 1. One right mandibular ramus 
with M , (TM M 4 1 1 72-302) and a left edentulous mandibular ramus (TM M 4 1 1 72-693) from 
level 1 of the hearth; right mandibular ramus with the incisor and P4 (TMM 41 172-724) from 
level 1 of blocks 1 3- 1 5; right mandibular ramus with the incisor (TM M 4 1 1 72-26) from level 3 
of blocks 22-24; left mandibular ramus with the incisor and P4 (TM M 4 1 1 72-524) from level 4 
of blocks 22-24 C. 

Description and Discussion. — The lower molars are not drawn out to points lingually as in 
Thomomys; there is a deep pit between the tooth row and the ascending ramus. The lower 
molars have enamel plates on posterior faces only. 

The plains pocket gopher is represented by a relatively small amount of material. It is usually 
found in areas of sandy soil more than 4 in. deep (Davis 1960) and is not recorded in Culberson 
County today. The nearest record is at Monahans, Ward County, Texas (Hall and Kelson 
1959). The closely related species G. arenarius is found no closer than El Paso County, Texas. 
The Pratt Cave specimens are almost exactly at the midpoint of the distributional gap between 
G. bursarius and G. arenarius, and may indicate that the separation of these two species is quite 
recent. The low frequency may indicate limited areas of appropriate soil with small populations 
in lowlands near the cave in the past. 

Thomomys bottae (Eydoux and Gervais), Botta Pocket Gopher 

Material.— Maxillaries and mandibles from blocks 2-4, level 2; blocks 5-7, level 2; blocks 8-9, 
levels 1, 2; blocks 10-12, levels 1, 2; blocks 13-15, levels 1, 2; blocks 19-21 A, level 3; blocks 
19-21 B, levels 2, 3; blocks 22-24 A, levels 2, 3; blocks 22-24 B, level 5; blocks 22-24 C, level 4; 



254 LUNDELIUS 

blocks 25-27 C, levels 3, 4, 5; level 1 of the hearth and the packrat nest of block 19. 
Description and Discussion. — The mandibles that are referred to this species cannot be dis- 
tinguished from those of modern representatives. The lower molars are drawn out to points on 
the lingual side. Some lower molars tend to be grooved on the labial side. The anterior lobe of 
P 4 is triangular. There is no deep pit between the tooth row and the ascending ramus of the jaw 
as in Geomys. The M 3 is subcircular. The incisive foramina open posterior to the anterior 
opening of the infraorbital canal. 

This is the most abundant gopher in the Pratt Cave deposits. It is common today in 
McKittrick Canyon close to Pratt Cave. 

Dipodomys ordii Woodhouse, Ord's Kangaroo Rat 

Material.— Left mandible from blocks 8-9, level 2 silt layer (TMM 41 172-688); right maxillary 
from blocks 10-12, level 2 (TMM 41 172-254); right mandible from blocks 13-15, level 1 (TMM 
41 172-71 1); left mandible from blocks 22-24 B, level 5 (TMM 41 172-658); left maxillary from 
blocks 25-27 B, level 4 (TMM 41172-664). 

Description and Discussion. — The masseteric foramen opens in a fossa which is steep-walled 
anteriorly as is Dipodomys spectabilis but differs from it in its smaller size. In D. merriami the 
masseteric foramen does not lie in a deep fossa but is in a shallow depression. The Pratt Cave 
specimen resembles D. ordii in the location of a small nutrient foramen close behind M 3 on the 
crest of the ridge extending posteriorly from M 3 . In D. merriami this foramen is located more 
posteriorly and is on the lingual side rather than on the crest of the ridge. 

The P 4 has enamel breaks at both ends of the hypolophid. The M 3 is oval and has enamel 
breaks at both lingual and labial ends. The enamel breaks develop late in this species (Wood 
1935). This specimen is apparently old, and the teeth are sufficiently worn to show the enamel 
breaks. The P 4 is roughly triangular, with the pattern worn away. The anterointernal corner has 
a distinct re-entrant, making the tooth asymmetrical. D. merriami lacks this re-entrant or has it 
very poorly developed. The M 1 is oval, with a re-entrant on the labial end. This is as found in 
Recent specimens of D. ordii. No enamel breaks are present in the Pratt Cave specimen, but 
according to Wood (1935), they appear late in life. In D. merriami the enamel breaks appear 
early. 

Dipodomys ordii is uncommon below the Guadalupe Escarpment today but is abundant else- 
where in Culberson County (Davis and Robertson 1944). 

Dipodomys spectabilis Merriam, Banner-tailed Kangaroo Rat 

Material.— Two left mandibles (TMM 41 172-683 and -712) from blocks 22-24 A and blocks 
10-12, level 1; and one right mandible (TMM 41172-271) from blocks 19-21 B, level 3. 
Description and Discussion. — The mandible resembles that of Dipodomys spectabilis in its size 
and robustness and in the presence of a broad, deep fossa on the lingual side of the ascending 
ramus of the mandible posterior to the M 3 . The P4 has the outline of a truncated triangle. The 
enamel is broken in four places — at the anterior corners and at the sides of the hypolophid. 
Similar breaks are present in the enamel of a Recent specimen of D. spectabilis (TNHC 642) 
from Presidio County, Texas. The Mi_2 are wider than the P 4 . All lower molars are less com- 
pressed anteroposteriorly than those of TNHC 642 and are more like those figured by Wood 
(1935: fig. 88). All lower molars have enamel breaks at both the lingual and labial ends as in the 
Presidio County specimen. 

According to Davis (1960), D. spectabilis is limited to low hills with sparse vegetation and 
hard, gravelly soil. Alluvial fans east of the Guadalupe Escarpment support large populations. 

Perognathus hispidus Baird, Hispid Pocket Mouse 

Material.— Abundant material from blocks 2-4, levels 2, 4; blocks 5-7, level 2; blocks 8-9, levels 
1,2; blocks 10-12, levels 1,2; blocks 13-15, levels 1, 2; blocks 19-21 A, levels 2, 3; blocks 19-21 
B, level 3, blocks 22-24 A, levels 2, 3; blocks 22-24 B, level 4; blocks 25-27 B, level 5; blocks 
25-27 C, levels 4, 5. 



POST-PLEISTOCENE MAMMALS 255 



Description and Discussion.— In size and morphology these specimens cannot be separated 
from Recent specimens of this species. All bones are too large to belong to any other species of 
Perognathus, which might be expected in this area. P. hispidus is found in areas of friable soil 
and moderate vegetation over wide areas of Texas and New Mexico, yet there is but one record 
for the southern Guadalupe Mountains (Davis 1940). 

Perognathus merriami Allen, Merriam's Pocket Mouse 

Material.— Mandibles from blocks 2-4, level 2; blocks 8-9, level 2; blocks 13-15, level 2; blocks 
16-18 B, level 1; blocks 19-21 A, levels 1,2; blocks 19-21 B, levels 3, 4; blocks 22-24 C, level 4. 
Description and Discussion. — The Pratt Cave material differs from Perognathus flavus in the 
greater posterior turning of the anterolabial cusps of Mi and M2, the lesser anteroposterior 
compression of the cross lophs of M 1 and M2 , the smaller size of the posterolabial cusps of Mi 
and M2 (especially M2), and the relatively deeper mandible. 

They differ from P. nelsoni in the relatively smaller P 4 , the much greater posterior extension 
of the anterolabial cusps of Mi and M2, and the much less parallel transverse lophs of teeth. 

They differ from P. penicillatus in the relatively smaller M3, lower crowned teeth, greater 
posterior development of the anterolabial cusp of M[ and M 2 , and a relatively shorter P 4 . 

They differ from P. flavescens in the larger P 4 and the presence of an external cingulum 
joining the two lophs on Mj and M2. 

P. merriami is widely distributed in western Texas, New Mexico, and northern Mexico; it 
occurs in areas of sandy or gravelly soils with sparse vegetation. Certainly it is present today in 
the area around Pratt Cave, but specimens are not available. The nearest record is the vicinity of 
Pine Springs (Davis and Robertson 1944). 

ORDER Lagomorpha 
Family Leporidae 

Lepus californicus Gray, Black-tailed Jackrabbit 

Material.— One left maxillary (TMM 41172-778) from blocks 16-18 A, level 1; one left 
mandible (TMM 41 172-834) from blocks 5-7, level 2; two upper molars (TMM 41 172-786 and 
-787) from blocks 2-4, level 2. Additional postcranial material not positively identifiable to 
species from blocks 2-4, level 2; blocks 8-9, levels 1 and silt layer of 2; blocks 10-12, levels 1,2; 
blocks 13-15, levels 1,2; blocks 19-21 A, level 3; blocks 19-21 B, level 2; blocks 22-24 A, levels 
2, 3; blocks 22-24 B, levels 2, 4; blocks 22-24 C, level 3; blocks 25-27 A, level 2; blocks 25-27 B, 
levels 3, 4, 5, 6; blocks 25-27 C, level 4; level 1 of the hearth; block 19, packrat nest. 
Description and Discussion. — The Pratt Cave material has been compared with Lepus 
americanus, L. townsendii, and L. californicus, and found to be indistinguishable from L. 
californicus. The inner and outer ends of the trigonids and talonids of the P 3 -M 3 are rounded as 
in L. californicus rather than pointed as in L. americanus. The labial groove separating the 
trigonid and talonid of the lower teeth is deep and parallel-sided as in L. californicus. This 
groove is shallow in L. americanus. The size is the same as that of Recent specimens of L. 
californicus and is larger than the specimen of L. americanus available for comparison. 

The upper molars of the Pratt Cave Lepus and Recent specimens of L. californicus are pro- 
portionately wider than those of L. townsendii and have the anterior and posterior ends less 
convex and more nearly parallel. The re-entrants of the upper teeth are narrower and extend 
closer to the labial margin of the teeth in L. californicus than in L. townsendii. The Pratt Cave 
Lepus material is similar to L. californicus in these characters. 

There is no evidence of the presence of either L. townsendii or L. americanus in the Pratt Cave 
material. Both of these species might be expected in deposits of late Pleistocene age in this 
region. 

L californicus presently is widely distributed in the lowlands below Pratt Cave. 

Sylvilagus audubonii (Baird), Desert Cottontail 

Material.— One right edentulous maxillary (TMM 41 172-36), one right maxillary with P 2 -M 3 
(TMM 41172-194), one left maxillary with P2 4 (TMM 41172-37) from levels 2 and 3. 



256 LUNDELIUS 

Description and Discussion. — Sylvilagus audubonii is so similar to S.floridanus in most cranial 
characters that it is extremely difficult to differentiate the two species. The palatine of S. 
audubonii has on its medial side a prominent rounded ridge which projects into the basi- 
pharyngeal canal. (This results in a rapid narrowing of the canal posteriorly.) The ridge is poorly 
developed in S.floridanus and the basipharyngeal canal narrows gradually posteriorly. The one 
specimen referred to S. audubonii shows the anterior part of this ridge on the medial surface of 
the palatine. 
This species is common around Pratt Cave. 

Sylvilagus sp. 

Material. — Abundant dental and postcranial material from all levels. 

Description and Discussion. — Most of the Sylvilagus material from Pratt Cave cannot be 
specifically identified. There is the possibility that some of it represents S.floridanus robustus, 
which is known from McKittrick Canyon today. 

LITERATURE CITED 

Ayer, M. Y. 1937. The archaeological and faunal material from Williams' Cave, 

Guadalupe Mountains, Texas. Proc. Acad. Nat. Sci. Phila. 88:599-618. 
Bailey, V. 1905. Biological survey of Texas. N. Am. Fauna 25:1-222. 
Bretz, J. H. 1949. Carlsbad Caverns and other caves of the Guadalupe Block, New 

Mexico. J. Geol. 57:447-463. 
Dalquest, W. W., E. Roth, and F. Judd. 1969. The mammal fauna of Schultze 

Cave, Edwards County, Texas. Bull. Fla. State Mus. Biol. Sci. 13:205-276. 
Davis, W. B. 1940. Mammals of the Guadalupe Mountains of western Texas. Oc- 

cas. Pap. Mus. Zool. La. State Univ. 7:69-84. 

1960. The mammals of Texas. Bull. Tex. Game Fish Comm. 2 7:1-252. 

Davis, W. B., and J. L. Robertson, J r. 1944. The mammals of Culberson County, 

Texas. J. Mammal. 25:254-27 3. 
Finley, R. B., Jr. 1958. The woodrats of Colorado, distribution and ecology. 

Univ. Kans. Publ. Mus. Nat. Hist. 10:215-552. 
Gaughran, G. R. L. 1954. A comparative study of the osteology and myology of 

the cranial and cervical regions of the shrew, Blarina brevicauda, and the mole, 

Scalopus aquaticus. Misc. Publ. Mus. Zool., Univ. Mich. 80:9-82. 
Gehlbach, F. R., and J. A. Holman. 1974. Paleoecology of amphibians and 

reptiles from Pratt Cave, Guadalupe Mountains National Park, Texas. South- 
west. Nat. 19:191-198. 
Guilday, J. P., P. S. Martin, and A. McCrady. 1964. New Paris No. 4: A 

Pleistocene cave deposit in Bedford County, Pennsylvania. Bull. Natl. Speleol. 

Soc. 26:121-194. 
Hall E. R., and K. Kelson. 1959. The mammals of North America. Ronald 

Press, New York. 
Harris, A. H. 1963a. Ecological distribution of some vertebrates in the San Juan 

Basin, New Mexico. Mus. New Mexico Pap. Anthropol. 8:1-64. 
1963b. Vertebrate remains and past environmental reconstruction in the 

Navajo Reservoir District. Mus. New Mexico Pap. Anthropol. 11:8-59. 
1970. The Dry Cave mammalian fauna and late pluvial conditions in south- 



eastern New Mexico. Tex. J. Sci. 22:3-27. 



POST-PLEISTOCENE MAMMALS 257 

Harris, A. H., and J. S. Findley. 1964. Pleistocene-Recent fauna of the Isleta 
Caves, Bernalillo County, New Mexico. Am. J. Sci. 262:114-120. 

Hibbard, C. W. 1960. An interpretation of Pliocene and Pleistocene climates in 
North America. Pages 5-30 in The President's Address, Rep. Mich. Acad. Sci. 
Arts and Letters, 1960 (for 1959-60). 

Hibbard, C. W., and D. W. Taylor. 1960. Two late Pleistocene faunas from 
southwestern Kansas. Contrib. Mus. Paleontol. Univ. Mich. 16:1-223. 

Hooper, E. T. 1952. A systematic review of the harvest mice (genus Reithro- 
dontomys) of Latin America. Misc. Publ. Mus. Zool. Univ. Mich. 77:1-225. 

Horberg, L. 1949. Geomorphic history of the Carlsbad Caverns area, New 
Mexico. J. Geol. 57:464-476. 

Lundelius, E. L., Jr. 1967. Late Pleistocene and Holocene faunal history of Cen- 
tral Texas. In P. S. Martin and H. W. Wright, eds. Pleistocene Extinctions: The 
Search for a Cause, Yale Univ. Press, New Haven. 

Martin, R. A. 1968. Further study of the Friesenhahn Cave Peromyscus. South- 
west. Nat. 13:253-266. 

Murray, K. F. 1957. Pleistocene climate and the fauna of Burnet Cave, New 
Mexico. Ecology 38:129-132. 

Park Staff. 1965. Mammals of Carlsbad Caverns National Park, New Mexico. 
Park circular, 10 pp. 

Schultz, C. B., and E. B. Howard. 1935. The fauna of Burnet Cave, Guadalupe 
Mountains, New Mexico. Proc. Acad. Nat. Sci. Phila. 87:273-298. 

Semken, H. A., Jr. 1961. Fossil vertebrates from Longhorn Cavern, Burnet 
County, Texas. Tex. J. Sci 13:290-310. 

Stearns, C. E. 1942. A fossil marmot from New Mexico and its climatic signifi- 
cance. Am. J. Sci. 240:867-878. 

Wood, A. E. 1935. Evolution and relationships of the heteromyid rodents with 
new forms from the Tertiary of western North America. Ann. Carnegie Mus. 
24:73-263. 



ACKNOWLEDGMENTS 

The author wishes to thank the following people for their assistance. The 
work was made possible by the various superintendents of the Carlsbad 
Caverns National Park. Park naturalists Ken Baker, Philip Van Cleave, and 
Peter Sanchez provided logistic and moral support. Rangers Roger Reisch, 
John Broadbent, and James Massey assisted in the field work. The following 
people provided assistance in the field and laboratory: Bill Balgemann, Roy 
Reeves III, John Greer, Mitchell Bronaugh, Ronnie Nims, Billy Davidson, 
Meredith Turner, Keith Bell, and Leon Long. Albert Schroeder coordinated 



258 LUNDELIUS 

the entire project by distributing information, material, and advice. Dr. 
Robert Hoffmann of the University of Kansas and Dr. James Findley of the 
University of New Mexico loaned comparative material. Victoria Hunter, 
Linda Young, and Jan Bannan aided in preparation of the manuscript. 
Financial support was provided by the Carlsbad Caverns Natural History 
Association, the Southwest Regional Office of the National Park Service, 
and the Hal P. Bybee Fund of the Geology Foundation of the University of 
Texas at Austin. 



Ground Sloth Dung of the 
Guadalupe Mountains 



W. GEOFFREY SPAULDING and PAUL S. MARTIN, 

University of Arizona, Tucson 

Eight caves in the North American continent are known to contain the 
dung of the extinct Shasta ground sloth (Nothrotheriops shastense). Four of 
these lie within the bounds of the Guadalupe Mountains National Park (Fig. 
1). The dried feces of this large extinct herbivore have been used to recon- 
struct its diet and the environment in which it lived (Laudermilk and Munz 
1934; Martin et al. 1961; Long etal. 1974). Through radiocarbon dating, the 
dung can be used to determine the time of occupation of the caves by the 
sloths. In this study, we compare Shasta ground sloth dung with that of other 
herbivores. We wish to learn more of the life history of this extinct species. 

The Upper Sloth Caves (Caves 05, 08, and 09; Logan and Black 1977) are 
located on the west-facing escarpment of the Guadalupe Mountains at 2000 
m elevation. They were first mentioned by Schultz (1943) who was present 
during the excavation of these caves in the 1930s. No bones identified as 
sloth have been recovered from test excavations. The dung recovered closely 
resembles that of sloths from better known deposits, such as Rampart Cave, 
Arizona. The stratigraphically controlled excavations of Logan in Cave 08 
provide most of the material for this study. Of particular interest to us is the 
sloth dung in Stratum 3, Trench 1, Cave 08. 

Williams Cave at 1500 m in the foothills southwest of the Guadalupe 
Mountains has been known for some time (Ayer 1936). Both cultural strata 
and underlying Pleistocene faunal deposits were removed during the course 
of archaeological excavations in the 1930s. Unfortunately, stratigraphic 
controls were imperfect. Dung identified by Laudermilk as that of the Shasta 
ground sloth was reported in Ayer (1936). 

METHODS 

Eleven ground sloth dung samples were subjected to fractional analysis. 
Six were from Stratum 3, Trench 1, Cave 08. Unprovenienced sloth dung 
from Williams Cave was obtained for study from the Carlsbad Public 
Library, Carlsbad, New Mexico, and the Academy of Natural Sciences, 

259 






260 



SPAULDING AND MARTIN 





Gypsum 
(£ave 








Rompart f 
/Cave- I 














X 1 ' 


r Muav Cave 




NEW MEXICO 










ARIZONA 




















£-Aden Crater 


,Cove 05 






/*~Cave 


Oft OQ 




N 










^Williams 
Cave 








i 






\ TEXAS 




1 




MEXICO 




% 











Fig. 1. Ground sloth dung caves of North America. 



TABLE 1. Type and provenience of dung samples used in fractional and pollen analyses. 



Sample no. 



Description 



1-6 
7-8 

9 
10-11 

12 

13 

14 

15 

16 

17-18 

19 

20 



Sloth. Cave 08, Trench 1, Stratum 3, ca. 35 cm. Guadalupe Mountains, 

Culberson County, Texas. 
Artiodactyl fecal pellets. Cave 08, Stratum 3, ca. 15 cm. Guadalupe 

Mountains, Culberson County, Texas. 
Cow. Collected in the vicinity of the Upper Sloth Caves, Guadalupe 

Mountains, Culberson County, Texas. 
Sloth. Unprovenienced dung balls from Williams Cave, Guadalupe 

Mountains, Texas. 
Cow. Collected in the vicinity of Rampart Cave, Grand Canyon, 

Arizona. 
Black rhinoceros. Serengeti Plain, vicinity of Olduvai Gorge, 

Tanzania. 
Wild burro. Warm Sulfur Springs, Panamint Valley, California. 
Cow. Medanos de Ricardo, Los Palomas, Argentina. 
Cow. Hanging Rock Cave, Churchill County, Nevada. 
Sloth. Unprovenienced dung balls from Rampart Cave, Grand Canyon, 

Arizona. 
Sloth. Unprovenienced dung ball from Gypsum Cave, Clark County, 

Nevada. 
Musk ox. Captive in the vicinity of Farmington, Vermont. 



SLOTH DUNG 



261 



-o 




G 


0) 


ad 


JD 


•s. 


F 


bi 


3 


ta 


z 


bj] 


a> 


c 


a 


t_l 


E 


* 


rt 


IMJ 


</5 


a 




£ 




>. 




~ 





r". 




e 




o 








o 




- 




f t-, 


a> 


E 


.a 

c/5 


c 


d 




a 


CTJ 


•_5 


- 


S 


Q 


Oh 




■a 


CS 


c 




rt 


W 




h- 1 


X, 


CQ 


&b 


^ 


a> 


H 


£ 



r- 


cs 


CS «o »o 

■<3- so r- 


r- io 


VO ^ 

oo r- 


O TT VD Tf 

ro — i cs r- 


st 


>o 


,-H r-H O 


O ro 


CS C> 


~H IO IO IO 

— H CS 



OO CS rf Tt 

— ( r- so C- 

r^ © «-h © 



^O M3 © 

o o r^ 



— i OO <^ ^h CS o 
© © © © © ro 

© © cs' co -^ cs" 



H fl Tt ^O V) t^ 

•* •* O M3 lO -h 

— J d -h © d a\ 



Tj- CO CS O O IO 

(N (^ \D Ifl >0 50 
^■' ^' (rj h rt 6 



CS — h ^ 



CO 00 — ( ^h 

0*\ CS O r-H 

O -*t CO* CS 



c-~ <j> o\ r-» ct\ 
o — h io so r-~ 



O <— I CO C- CS ^H 

t^ VO lO -H M f 

ro' — ' CS* — h r-H On 



c 
















a) 
g 




















a 


E 


E 


E 




OJ 






g 


E 


E 


E 




Qj 




CO 


© 

cs 


© 


© 




r° 






V 


V 


V 


V 




<D 


Oh 














43 
















, — s 


'+-i 














W) 


o 


£ 


E 


g 


P 


g 


g 


,13 




s 


g 


E 


E 


E 


g 


T 


o 


o 


so 


CO 


co 


















<D 


0) 


CO 


rsi 


— 


© 


© 


© 


^ 


^ 


A 


A 


A 


A 


A 


V 



Tfr o CS vo CS 
OS IO OO Tf O 



to 


CS 


SO 


n 


CS 


- 


X 

SO 


SO 
CO 


X 


C 


IO 


© 
o> 


CO 


CS 


r~ 


-t 


■<fr 


so 


CS 

q 


OS 


CO 

© 


CS 


30 
SO 


00 



© © in r— as r-- 



IO 

SO 


SO 

(-1 


CO 

X 


© 

O 


X 


X 

o 


3 


ro 


r- 


>o 


-r 


so 


cs 

er- 


CO 
CO 


cs 


CO 

CO 


ro 
CO 


CO 

X 


r- 


X 


so 


o 


VO 


«o 


so 


r- 

© 


00 
CO 


CO 


o> 

CS 


CC 

-r 


r-~ 


CO 


SO 


co 


SO 


o> 

so 


»o 
io 


OO 




IO 


1- 


SO 

o 


00 
CS 


^ 


oc 


SO 


-t 





Cf\ so © ro O^ co 
CS vo o> o> ^^ © 



oc 


CS 


o> 


Tt 


SO 


0> 




-^r 


© 
»o 


a-. 


CO 


oo 
cs 


CS 


»o 


~/C 


CO 


T 


>o 

so 



E E E E 

SEES 
-t o o VO 

ro cs ^-1 d 
V V V V 



E E E E E E 

E E E E E E 

rj- o © so co ro 

ro CS ^h O © © 

A A A A A V 



262 



SPAULDING AND MARTIN 



© 


in 
It 


rs 

o 


ON 
rS 


co 


cc 
in 


o 
© 


cs 


O 
CN 




r^ 


© 


in 




CM 


o 


o 


O 


r] 


" 


rj 


vc 


O 


CS 


rS 

rs 


in 


o- 


UO 


0> 




CO 

ro 


ON 

in 




rs 


O 

in 


ON 
ON 


rS 


i- 


rS 




00 


r i 
rs 




VD 


O 


o 


O 


o 


o 


co 


in 


on 


CN 


CO 


o 


rs 



-i 


.a 


a> 


1/2 


3 


CU 


4J 


t) 


(3 




2 


ft 




£ 




"O 


rs 


c 




5 


— 




mJ 


•£. 


pq 


bb 


" 


V 


H 


£ 



— i oo rs io r- r- 
rs on co rs oo rs 

d 



^o «-h co oo in oo 
rs oo oo Tf —■ tj- 



© 


,_H 


d 


d 


CO 


T_l 


in 


c-^ 


** 


in 


csd 


<* 


CO 


o 
to 


cc 


ON 


ON 


in 


o 


c 

CO 




co 


rS 


— 


o 


© 


r~- 


CO 


t 


— 


CO 


in 


r- 1 



i— i rs r- m >n on 
q co oq on no on 
d d rs rs' Tf io 



O H O *H 



«-H — in 



ON 




-r 


rs 


CO 


o 

ON 


co 


ON 


rs 


in 

ON 


- 
ro 


© 


© 


" 


rs 


cs 


rs 


rs 


"* 


in 


ON 


ON 


rs 


in 




CO 

rs 


rs 


so 


in 


o 

00 


o 


NO 

CO 


NO 
© 


rs 


in 

t 


3" 


rs 


«-" 


cs 


•- 1 


"- 1 


CO 


rs 


ON 


r- 


rs 


^ 


00 

r j 



rt 






E 


£ 


E 


E 








E 


E 


E 


E 




(U 






£ 


g 


E 


E 








E 


£ 


£ 


£ 




*-' 


C/2 




rf 


q 


q 


NO 




u 




-r 


© 


q 


NO 


































j_ 


Cj 




CO 


rs" 


i — 


d 




CJ 




CO 


rs 


i — i 


d 




o 






V 


V 


V 


V 




'■£ 




V 


V 


V 


V 




a> 


>— 














rd 














ja 


a 














Oh 














^ac 


o 


£ 


g 


g 


g 


E 


£ 


O 


E 


E 


P 


g 


E 


E 


g 




E 


E 


E 


E 


E 


£ 


■£2 


g 


g 


£ 


£ 


^ 


£ 


.5? 


rt 


o 


q 


VO 


m 


CO 




■* 


q 


q 


v£> 


CO 


CO 


'5 


'S 


CO 


rs 


,-H 


d 


© 


d 


a. 


CO 


rs 


_H 


d 


- 


d 


£ 


£ 


A 


A 


A 


A 


A 


V 


A 


A 


A 


A 


A 


V 



SLOTH DUNG 



263 







COW 


°1 




— 9 






15 

16 


01 


.^=_=_ 


::^^^ 



3.4 2.0 10 0.6 0.3 



SLOTH 




3.4 2.0 1.0 0.6 0.3 





OTHERS 




40n 




13 






14 

20 


20- 


i i 


~- --._ 




— i 1 



3.4 2.0 1.0 0.6 0.3 

Mesh size in mm. 



Fig. 2. Graphic representation of dung size-frequency distribution derived from 
fractional analysis. See Table 1 for an explanation of sample numbers. 



Philadelphia, Pennsylvania. Fossil ground sloth dung from Rampart Cave, 
Grand Canyon, Arizona; and Gypsum Cave, Nevada; and modern dung 
samples of cow, burro, rhinoceros, and musk ox were used for comparison 
(Table 1). 

Each dung sample was dried, weighed, and disaggregated in a solution of 
water and detergent. Sonic vibration proved effective in separating tightly 
compacted fecal material. When disaggregation was complete, the mixture 
was decanted through a series of five sieves of decreasing mesh size — 3.4 mm, 
2.0 mm, 1 .0 mm, 0.6 mm, and 0.3 mm. The initial decant passing through the 
sieves was captured and used for pollen analysis. Fecal material from each 
sieve was removed, dried, and weighed separately. The weight of solubles 
and particles less than 0.3 mm in size was obtained by adding the weight of 



264 



SPAULDING AND MARTIN 



TABLE 3. Pollen counts of fecal samples from Cave 08 and Williams Cave. See Table 1 for type 
and provenience of samples. 











Sample nc 


>. 










Pollen types 


1 


2 


3 


4 


5 


6 


7 


8 


10 


11 


Picea 


4 


6 


1 






1 










Pseudotsuga 






2 
















Pinus 


97 


16 


30 


24 




15 


13 


11 


9 


1 


Juniper us 


2 


1 


9 


3 


2 


4 


3 


3 


6 


16 


Ostrya 


4 


2 


16 


1 




5 


11 


1 


1 




Quercus 


4 




2 








2 


2 


3 


2 


Acer 


2 










1 










Celtis 












3 










Cheno-am 




1 








2 


1 


1 


10 


57 


Gramineae 


15 


37 


8 


6 


19 


34 


73 


135 


16 


20 


Artemisia 


24 


81 


26 


116 


46 


65 


31 


4 


78 


20 


Short-spine Compositae 


2 


6 


2 


8 


10 


4 


2 


2 


2 


12 


Long-spine Compositae 


29 


40 


95 


30 


110 


45 


43 


21 


54 


50 


Senecio-Vype 






















Cercocarpus-type 


2 


2 


3 


1 


2 






1 


2 


3 


Petrophy turn-type 
















11 






Prunus-type 


1 




















Potentilla-type 












1 










Scrophulariaceae 








1 










1 


1 


Onagraceae 








1 














Leguminosae 
















1 


3 




Saxifragaceae 




1 




1 


4 


1 




2 


2 




Rhus 




1 














4 




Cylindropuntia 






1 
















Platyopuntia 










1 


1 










Caryophyllaceae 












1 










Euphorbia 














1 






4 


Malvaceae 




















1 


Boraginaceae 








3 


1 












Polemoniaceae 




1 














1 




Arceuthobium 


1 




1 


2 














Rubiaceae 




1 






1 


1 


3 








Berberis 


















1 




Rhamnus 




















1 


Solanaceae 












1 










Kallstroemia 




















1 


Larrea-type 














5 








Sparganium- Typha 














1 








Umbelliferae 














1 








Undetermined 


3 


1 


2 




1 


1 






2 


1 


Unknown 


10 


3 


2 


3 


3 


14 


9 


1 


5 


10 


N 


200 


200 


200 


200 


200 


200 


200 


200 


200 


200 


Sap 


113 


25 


60 


28 


2 


29 


29 


16 


19 


19 


Trilete, psilate spores 






1 


3 














Grains per gram(*10 3 ) 


29.9 


31.1 


11.6 




112.6 


6.0 


47.9 


124.0 


59.3 





SLOTH DUNG 265 

plant fragments captured in each sieve and subtracting that from the initial 
dry weight of the sample. Relative percentages calculated from the initial dry 
weight were used in comparing the samples (Table 2). 

Separate pieces of samples 1, 3, 10, and 11 were submitted to the 
Laboratory of Isotope Geochemistry, Department of Geosciences, Univer- 
sity of Arizona, for radiocarbon analysis. An additional sloth dung sample 
from a separate area in Cave 08, one from Cave 05, and artiodactyl fecal 
pellets from Stratum 3, Cave 08 were also dated (Table 3). 

Decant recovered from the screening process was centrifuged. The 
remaining solids were extracted for pollen following the methods for fecal 
material outlined by Spaulding (1974). Before fractional analysis, a known 
amount of Eucalyptus pollen grains was added to each sample to enable us to 
determine the amount of fossil pollen per gram of dung (Stockmarr 1973). 
Two hundred pollen grains were counted in each sample and relative fre- 
quencies of each pollen type were calculated from this pollen sum (Martin 
1963). Eucalyptus pollen and spores were counted separately and not 
included in the pollen sum (Table 3). 

RESULTS 

Fractional separation of herbivore dung produces a size-frequency curve 
with trends that appear to be unique to the dung of particular species (Table 
2, Fig. 2). The dung of ruminants, such as cow or musk ox, contains high 
percentages of fine particles. A nonruminant browser, such as rhinoceros, 
has very coarse dung. The size-frequency curve of burro dung is most similar 
to that of sloth, but has a greater percentage of fine particles. Differences in 
the size-frequency curves are greater between than within a species. Diets of 
the sloths that occupied Williams Cave and those that occupied Cave 08 
differed markedly. Dung from the former site consists mostly of Yucca fiber, 
whereas dung from the latter contains mostly twigs. Size-frequency curves 
produced by the dung from Williams Cave and Cave 08 are quite similar 
(Fig. 2). 

Analysis of pollen from sloth dung shows that most samples have high 
relative percentages of long-spine Compositae and Artemisia pollen (Table 
3). Relative percentages of arboreal pollen (AP), particularly pine (Pinus), 
are higher in the sloth dung samples from Cave 08 than from Williams Cave. 
Four-wing salt-bush (Atriplex canescens) seeds were found in the Cave 08 
sloth dung. Cheno-am (Chenopodiaceae plus Amaranthus) pollen is infre- 
quent in those samples. However, cheno-am pollen is relatively abundant in 
the samples from Williams Cave. 

The amount of Eucalyptus pollen counted with every 200 fossil pollen 
grains varies greatly. Determinations of the amount of fossil pollen per gram 
of sloth dung vary from 112,600 to 6000 grains per gram. It appears that 
much fossil and Eucalyptus pollen was trapped in the screens during 



266 SPAULDING AND MARTIN 

fractional analysis and was therefore not included in the decant used for 
pollen analysis. Hence, values for absolute amounts of pollen are con- 
sidered to be unreliable and are not used. In future studies, a separate sec- 
tion of dung should be reserved for pollen analysis. 

In addition to sloth dung, fossil artiodactyl fecal peilets from Stratum 3, 
Cave 08 were analyzed for pollen. These pellets are much larger than those 
dropped by deer (Odocoileus), sheep (Ovis), or modern mountain goat 
(Oreamnos americanus). They are closer in size and weight to pellets found 
in Rampart and Stanton's caves, Grand Canyon, Arizona, which were 
assigned by Iberall (1972) to the extinct Harrington's mountain goat 
(Oreamnos harringtoni). They also resemble the fecal pellets of the extant 
elk (Cervus canadensis). Bones of Oreamnos have been recovered from 
Pleistocene-age deposits in the Guadalupe Mountains (Logan, pers. comm.) 
and Merriam's elk (Cervus canadensis merriami) was present in the 
Guadalupes until Recent times. A radiocarbon date on the fossil pellets from 
Stratum 3, Cave 08 yielded an age of 1 1,760 ± 610 BP (radiocarbon years 
before present). 

DISCUSSION 

Through fractional analysis we have been able to discriminate among the 
dung of five herbivore genera. Although our sample is small, the data indi- 
cate that ground sloth dung may be quantitatively identified without having 
to rely upon associated faunal remains. Variation in the relative percentages 
of particles greater than 2.0 mm in size are attributed to dietary differences. 
Sample 18 from Rampart Cave is remarkably fine textured for ground sloth 
dung (Table 4). Before analysis, it was noted that this sample was degraded, 



TABLE 4. Radiocarbon ages of sloth dung from the Guadalupe Mountains. All age determina- 
tions are by the Laboratory of Isotope Geochemistry, Department of Geosciences, Univer- 
sity of Arizona and are given as radiocarbon years before the present (1950; BP). 

Sample Radiocarbon 

number number Description Age (BP) 

1 A-1583 Sloth dung. Cave 08, Trench 1, Stratum 3, 10,750±140 

ca. 30 cm. 

2 A- 1584 Sloth dung. Cave 08, Trench 1, Stratum 3, 1 1,060 ±180 

ca. 30 cm. 

Sloth dung. Williams Cave. No provenience. 11,93Q±170 

Sloth dung. Williams Cave. No provenience. 11, 140 ±320 

Sloth dung. Cave 05. No provenience. 11, 590 ±230 

Artiodactyl fecal pellets. Cave 08, Stratum 3, 1 1,760±610 

ca. 15 cm. 

Sloth dung. Cave 08, Stratum 3, ca. 15 cm. 10,780 ±140 

Sloth dung. Williams Cave. No provenience. 1 2, 100 ±210 



11 


A- 1588 


10 


A- 1589 


- 


A 1519 


- 


A-1533 


_ 


A- 1534 


- 


A- 1563 



SLOTH DUNG 267 

perhaps by fungi. It is unlike well-preserved ground sloth dung from the 
cave. 

The plant communities existing in the area at the time the sloths occupied 
these caves differed markedly from those of today. Needles of Douglas fir 
{Pseudotsuga menzeisii) and southwestern white pine (Pinus strobiformis) 
were found in the dung balls from Cave 08. These macrofossils and the radio- 
carbon dates (Table 4) indicate that the Upper Sloth Caves were frequented 
during the biochronological zone designated Wisconsin 2 (W2) by Van 
Devender et al. ( 1977). The presence of spruce (Picea) pollen in samples 1 , 2, 
3, and 6 supports their assertion that, although spruce is no longer present in 
the macrofossil record, it had not been completely extirpated from the 
Guadalupe Mountains by ca. 1 1,000 BP. Radiocarbon ages of the Williams 
Cave sloth dung (Table 3) indicate that this site may have been frequented by 
Nothrotheriops during zone Wl (Van Devender et al. 1976). 

The clustering of radiocarbon dates and the relatively thin layer of sloth 
dung found in Cave 08, Stratum 3 suggests that sloths occupied this cave 
only briefly. The deepest known deposit of ground sloth dung, 1 .4 m in depth 
representing over 30,000 years of sporadic occupation, comes from Rampart 
Cave, Arizona. Like the sloths of Rampart Cave, those of the Guadalupe 
Mountains apparently did not frequent the caves during the Wisconsin 
maximum between ca. 24,000 and ca. 14,000 BP. At present, there is no 
established radiocarbon record of ground sloths in any cave during the full- 
glacial. Older records (pre-Wisconsin maximum) of sloths in the Guadalupe 
Mountains have not been established, possibly due to the poor preservation 
of organic matter in the deeper strata of the Upper Sloth Caves. Sloths 
occupied the Upper Sloth Caves coincidentally close to the time of their ulti- 
mate demise, which Long et al. (1974) place at ca. 11,000 BP. 

The absence of Nothrotheriops from Rampart Cave fromca. 24,000 to ca. 
14,000 BP cannot necessarily be attributed to unfavorable environmental 
factors. During the occupation of the Upper Sloth Caves, the vegetation that 
prevailed was typical of much higher and cooler elevations than would have 
surrounded Rampart Cave even during the full-glacial (Phillips and Van 
Devender 1974). Similarly, Shasta ground sloth extinction in the Guada- 
lupe Mountains cannot easily be attributed to changing climate. Sloths lived 
in Gypsum Cave in southern Nevada at a time when the late Pleistocene cli- 
mate supported Joshua-tree vegetation at 610 m (Wells and Berger 1967). 
The average annual precipitation and temperature needed to support such a 
community implies that it was probably as xeric or even more xeric than 
present conditions in the vicinity of the Guadalupe Mountains sloth caves. 

Radiocarbon ages of the Guadalupe Mountains sloth dung are not 
appreciably younger than the time of Clovis big game hunters associated 
with mammoths which Haynes (1971) puts at 11,200 years ago. Possible 
exceptions are the two samples from Cave 08 (Table 4). Contamination by 
younger material cannot be ruled out as the cause of these younger dates. 



268 SPAULDING AND MARTIN 

Further effort is needed to replicate them. Possibly ground sloths lingered 
slightly longer in the Guadalupes than elsewhere. 

CONCLUSIONS 

Fractional analysis provides a quantitative basis for distinguishing fossil 
sloth dung from that of other herbivores. 

The late Pleistocene distribution of the Shasta ground sloth ranges from 
extremely xeric sites such as Gypsum Cave, Nevada, at 610 m to the semi- 
xeric Upper Sloth Caves of the Guadalupe Mountains at 2000 m. Wide eco- 
logical amplitude argues against any megafaunal extinction hypothesis 
founded upon climatic change. 

Qualitative examination of Shasta ground sloth dung from Williams Cave 
and the Upper Sloth Caves shows considerable difference in composition. 
This suggests a high degree of versatility in the sloths' dietary preferences. 

Results of radiocarbon analysis of dung from the Guadalupe Mountains 
sloth caves suggest that the animals might have lived a few hundred years 
longer than Long et al. (1974) suggested on the basis of dung deposits else- 
where. 

LITERATURE CITED 

Ayer, M. Y. 1936. The archaeological and faunal material from Wiliiams Cave. 

Guadalupe Mountains, Texas. Proc. Acad. Nat. Sci. Phila. 88:5899-619. 
Haynes, C. V. 1971. Geochronology of man-mammoth sites and their bearing on 

the origin of the Leano. Pages 77-92 in W. Dort, Jr., and J. K. Jones, Jr., eds., 

Pleistocene and Recent Environments of the Central Great Plains. 
Iberall, E. R. 1972. Paleoecological studies from fecal pellets: Stanton's Cave, 

Grand Canyon, Arizona. M.S. thesis, Univ. Arizona, Tucson. 
Laudermilk, J. D., and P. A. Munz. 1934. Plants in the dung of Nothro- 

therium from Gypsum cave, Nevada. Carnegie Inst. Wash. Publ. 453:29-37. 
Logan, L. E., and C. C. Black. 1977. The Quaternary vertebrate fauna of Upper 

Sloth Cave, Guadalupe Mountains National Park, Texas. This volume. 
Long, A., R. M. Hansen, and P. S. Martin. 1974. Extinction of the Shasta 

ground sloth Cave. Bull. Geol. Soc. Am. 85:1843-1848. 
Martin, P. S. 1963. The last 10,000 years: a fossil pollen record of the American 

southwest. Univ. Arizona Press, Tucson, 87 pp. 
Martin, P. S., B. E. Shutler, Jr., 1961. Rampart Cave coprolite and ecology of 

the Shasta ground sloth. Am. J. Sci. 259:102-127. 
Phillips, A. M., Ill, and T. R. Van Devender. 1974. Pleistocene packrat 

middens from the Lower Grand Canyon of Arizona. J. Ariz. Acad. Sci. 

9(3):117-119. 
SCHULTZ, C. B. 1943. Some artifact sites of early man in the Great Plains and adja- 
cent areas. Am. Antiquity 8:242-249. 
SPAULDING, W. G. 1974. Pollen analysis of fossil dung of Ovis canadensis from 

southern Nevada. Unpublished M.S. thesis, Univ. Arizona, Tucson. 
Stockmarr, J. 1973. Tablets with spores used in absolute pollen analysis. Pollen 

Spores 13:615-621. 



SLOTH DUNG 



269 



Van Devender, T. R., W. G. Spaulding, and A. M. Phillips, III. 1977. Late 
Pleistocene plant communities in the Guadalupe Mountains, Culberson County, 
Texas. This volume. 

Wells, P. V., and R. BERGER. 1967. Late Pleistocene history of coniferous wood- 
land in the Mohave Desert. Science 155:1640-1647. 



ACKNOWLEDGMENTS 

We thank Lloyd E. Logan, Texas Tech University, for assistance in the 
field and for providing most of the material for this study. Thomas R. Van 
Devender, Arthur M. Phillips, III, and Jim I. Mead, University of Arizona, 
Tony L. Burgess, Texas Tech University, and Benjamin and Cindi Everitt, El 
Paso Archaeological Society, helped with the field work. We received gener- 
ous help from Roger Reisch, Gary Ahlstrand, Phil Van Cleave, and John 
Chapman of the National Park Service. The Carlsbad Public Library, 
Carsbad, New Mexico, and the Academy of Natural Sciences of 
Philadelphia, Pennsylvania, provided fossil material for analysis. Betty Fink 
aided in editing and typing the manuscript. Austin Long and Paul Damon, 
Laboratory of Isotope Geochemistry, Department of Geosciences, Univer- 
sity of Arizona, provided the radiocarbon dates with the support of NSF 
Grant GB-27406 to Paul S. Martin. 

Contribution no. 749, Department of Geosciences, University of Arizona. 



Mammals of the Guadalupe Mountains 
National Park, Texas 



HUGH H. GENOWAYS, ROBERT J. BAKER and JOHN 
E. CORNELY, Texas Tech University, Lubbock 

The Guadalupe Mountains National Park was authorized by an act of 
Congress on 15 October 1966 and was formally established on 30 September 
1972. The park covers 76,468.6 acres located in Culberson and Hudspeth 
counties of Trans-Pecos Texas. The park contains the Texas portion of the 
uplifted Capitan Reef of Permian age. The southern end of the escarpment is 
marked by the prominent El Capitan. The escarpment extending northwest 
from El Capitan contains other impressive peaks including Guadalupe Peak, 
which at 8759 ft is the highest point in Texas. 

The low and intermediate elevations in the park contain floral and faunal 
elements from the Chihuahuan Desert. The high elevations are inhabited by 
montane elements with Rocky Mountain affinities. These montane ele- 
ments represent an island surrounded by, and in dynamic equilibrium with, 
the desert flora and fauna. The mountains, all canyons, and desert areas con- 
tain many fragile floral and faunal microhabitats. To preserve the natural 
heritage of the park, baseline data are being gathered by the National Park 
Service for use in development of the park's master plan. 

The first mammal survey of the Guadalupe Mountains was conducted by 
Vernon Bailey of the U.S. Biological Survey between 9 and 25 August 1901 
(Bailey 1905). His field notes and specimens are deposited in the National 
Museum of Natural History. During his visit,Bailey worked in Upper Dog 
Canyon, McKittrick Canyon, and various portions of the high country. 
Bailey (1905) reported 17 species inhabiting the Guadalupe Mountains and 
listed two additional species that possibly occurred there. The next work in 
the area was conducted by William B. Davis and field parties from Texas A 
& M University during 1938, 1939, and 1940 (Davis 1940; Davis and 
Robertson 1944). They worked at seven stations including McKittrick 
Canyon, West Dog Canyon, The Bowl, Burned Cabin, Pine Springs and 
Bear Canyons, Frijole, and 7 miles N Pine Springs. A total of 35 species of 
mammals (Davis 1940; Davis and Robertson 1944) were recorded as 
occurring in the Guadalupe Mountains as a result of this survey. LaVal 

271 



272 



GENOWAYS ET AL. 



(1973) studied the distribution and ecology of bats in McKittrick Canyon 
during 1968 and 1970. He presented data on 13 species. 

Our survey began in late May 1973 and continued through August 1975. 
The objectives of our study, which was supported by the National Park 
Service, were to survey the mammals occurring in the Guadalupe Mountains 
National Park, Texas, and to correlate their distribution with major plant 
associations. This inventory of the natural resources of the park is pre- 
liminary to the development of any serious management program. There- 
fore, we present the following accounts to serve as baseline data for future 
mammalian work in the park and development of the master plan for the 
park. 



»57 



58* 



•59 



! 



•17 






28 



4 < 



\r-V 



\ 

\ -7 



x 16i 



&* 



V 



•74 ; 

• 6i "SI 5 1 



• 3 



s 



1{ 



7 



si 



26* 



7^7\ 



•27 



<&--y 



■ 






71* *70 

7 ?\ ••48,49! 30*' 
• 56 73 T**\ ©53,55 ;1 

52 54 >50,51 
64* • M _^_ | 
68* ^ 65 ,66 » 6 7 ; 

63* I n r 

.6? # _ 






J 




\9" \ f . * 



r 



•1 

•24 



5«/.*6 
-•4 



A- 



P. V. 2 

22 >^ 

\^— 32 

- 38 #39^1 #35,36 
37^0*41,42 

4344 45* • 

47» <46 



Fig. 1. Map showing collecting localities in the Guadalupe Mountains National 
Park, Texas. Numbers correspond to those given in text with the exact location of 
each place. 



MAMMALS 273 







lis 






:«,:#%-; 








• 



■< 



Fig. 2. {Upper) Photograph of eastern slope of Guadalupe escarpment showing 
Frijole and Manzanita Spring. 

Fig. 3. {Lower) Photograph of a dry arroyo in McKittrick Canyon in the Guada- 
lupe Mountains National Park, Texas. 



274 



GENOWAYS ET AL. 



METHODS AND MATERIALS 

During our survey of mammals of the Guadalupe Mountains National 
Park in 1973-75, we visited selected sites throughout the park (see below) 
during all seasons of the year. Rodents were collected using various types of 
traps including museum specials, Sherman live-traps, Victor steel traps, and 




. . • ■ :■ .. ... ' ■■ 








- .' 'iJffr 



Fig. 4. {Upper) Photograph of southwestern face of El Capitan {right) and 
Guadalupe Peak {center) showing creosote bush community typical of the western 
and southern lowlands in the Guadalupe Mountains National Park, Texas. 



Fig. 5. {Lower) Photograph of the coniferous forest in The Bowl area in the Guada- 
lupe Mountains National Park, Texas. 



MAMMALS 275 

National live traps. Bats were obtained by mist-netting, by shooting 
individuals as they flew at dusk, and by inspecting daytime roosts. Carni- 
vores were taken by shooting and trapping and rabbits were obtained by 
shooting. All individuals taken in our work were prepared as various types of 
standard museum specimens. These specimens and extensive field notes 
made during our work are deposited in The Museum of Texas Tech Univer- 
sity (TTU). 

In addition to our material, we have examined specimens (abbreviations 
used to identify specimens in text) deposited in the Texas Cooperative Wild- 
life Collection, Texas A & M University (TCWC) and National Museum of 
Natural History (USNM), Washington, D.C. All cranial measurements 
were taken by means of dial calipers; external measurements were those 
recorded by the field collector. All measurements are recorded in milli- 
meters. Specimens were karyotyped using the methods of Baker (1970). 

COLLECTING LOCALITIES 

Listed below are the collecting localities visited during our survey of the mammals of the 
park. Locality numbers correspond to those given in Fig. 1. 

Following each locality or groups of localities below is a brief description of the habitat being 
sampled. Four major habitats in which work was conducted are shown in Figs. 2-5. Comely 
(1976) has presented a checklist of the mammals of the park with their major habitat preference. 

Culberson County 

1) Bear Canyon-Pump House (el. 1829 m). — The ruins of a pump house containing very 
large water pumps are situated on the Bear Canyon trail above upper Pine Spring. The vegeta- 
tion is open canyon woodland including Quercus grisea, Juniperus pinchotii, Arbutus 
xalapensis, Bouteloua gracilis, and Bouteloua curtipendula. 

2) Blue Ridge Campground (el. 2438 m).— Blue Ridge Campsite is situated at the north end 
of Blue Ridge which extends due north of Bush Mountain. The vegetation is open woodland in- 
cluding Pinus ponderosa, Quercus gambelii, Pseudotsuga menziesii, Juniperus deppeana, 
Bouteloua gracilis, Muhlenbergia pauciflora, and M. dubia. 

3) Bush Mountain (el. 2530 m).— Bush Mountain forms part of the western ridge of the 
Guadalupe Mountains. Traps were set in pine-oak meadow immediately southeast of the sum- 
mit. Plants included Pinus ponderosa, Quercus gambelii, Pseudotsuga menziesii, Juniperus 
deppeana, Ceanothus greggii, Cercocarpus montanus, Bouteloua gracilis, Muhlenbergia 
pauciflora, M. dubia, and Hymenoxys richardsonii. 

4) Frijole Ranger Station (el. 1692 m); 5) Manzanita Spring (el. 1676 m); 6) Nipple Hill (el. 
1646 m).— Frijole Ranger Station, Manzanita Spring, and Nipple Hill are on the bajada east of 
the Guadalupe Escarpment. The vegetation is open juniper woodland and grassland including 
Juniperus pinchotii, Muhlenbergia set if olia, Bouteloua gracilis, B. warnockii, and Parthenium 
incanum. 

7) Grisham-Hunter Lodge (el. 1615 m); 8) Vi mi. NNE Grisham-Hunter Lodge (el. 1615 m); 
9) Half-way between Pratt Lodge and McKittrick Canyon Parking Lot (el. 1554 m); 10) North 
McKittrick Canyon at Devils Den Canyon (el. 1585 m); 11) Pratt Lodge (el. 1585 m); 12) 0.3 mi. 
N, 0.5 mi. E Pratt Lodge (el 1570 m); 13) Stone Cabin above Grisham-Hunter Lodge (el. 1645 
m); 14) Thrush Hollow, »/4 mi. S Pratt Lodge (el. 1 590 m).— The vegetation of the canyon walls 
is succulent desert, whereas the canyon floor vegetation is canyon woodland. An intermittent 
stream in McKittrick Canyon is the only permanent stream in Guadalupe Mountains National 
Park. Plants in the canyon include Acer grandidentatum, Quercus muhlenbergii, Arbutus 



276 GENOWAYS ET AL. 

xalapensis, Pinus ponderosa, Stipa tenuissima, Muhlenbergia emersleyi, Quercus undulata, 
Juniperus deppeana, and Dasylirion leiophyllum. 

15) Guadalupe Peak Campsite (el. 2439 m). — Guadalupe Peak Campsite is situated on top of 
the eastern escarpment due east of Guadalupe Peak. The vegetation is open coniferous wood- 
land dominated by Pinus ponderosa and Muhlenbergia pauciflora. 

1 6) Lost Peak Mine (el. 2 1 64 m). — Lost Peak Mine is an old copper mine between Upper Dog 
Canyon Ranger Station and Lost Peak. The mine is on the west slope of the ridge which the trail 
from the ranger station to Lost Peak traverses. The vegetation is chaparral and succulent desert 
including Ceanothus greggii, Cercocarpus montanus, Nolina micrantha, Dasylirion 
leiophyllum, and Quercus undulata. 

17) Marcus Cabin-West Dog Canyon, 6 3 / 8 mi. N, 3 / 4 mi. W Guadalupe Peak (el. 1905 m).— 
West Dog Canyon is relatively large, with a steep wall forming the east side and a more gradual 
slope marking the west side. Deep soil of the canyon floor is cut by dry washes. The vegetation is 
mixed grassland with riparian vegetation along the washes. Plants include Muhlenbergia 
repens, Bouteloua gracilis, B. warnockii, Aristida glauca, Berberis haematocarpa, Fallugia 
paradoxa, Opuntia imbricata, Xanthocephalum sarothrae, Xanthium spinosum, Verbesina 
encelioides, and Pinus edulis. 

18) McKittrick Canyon Parking Lot (el. 1524 m). — The parking lot is on the canyon floor at 
the mouth of the canyon. Canyon floor vegetation is open grassland with succulent desert on the 
slopes. Species include Agave lecheguilla, Dasylirion leiophyllum, Bouteloua eriopoda, B. 
gracilis, Muhlenbergia setifolia, and Juniperus pinchotii. 

19) Mescalero Campground, A X A mi. N, % mi. E Guadalupe Peak (el. 2286 m). — Mescalero is 
situated on top of a ridge separating West Dog Canyon drainage and South McKittrick Canyon 
drainage. The campsite is on the trail between Upper Dog Canyon Ranger Station and The 
Bowl. The vegetation is woodland including Pinus edulis, P. ponderosa, Quercus undulata, 
Juniperus deppeana, Nolina micrantha, Ceanothus greggii, Rosa stellata, and Muhlenbergia 
dubia. 

20) Patterson Hiils Notch, 3 1 / 1 6mi. S, 1 3 / 8 mi. W Guadalupe Peak (el. 1 356 m); 2 1 ) 2>V% mi. S, 
VA mi. W Guadalupe Peak (el. 1341 m); 22) 3»/ 8 mi. S, l 3 / 8 mi. W Guadalupe Peak (el. 1356 m); 
23) 3V4 mi. S, 2 3 / 8 mi. W Guadalupe Peak (el. 1 341 m). — Water erosion has cut the notch through 
the eastern ridge of the Patterson Hills along the Williams Ranch Road. The vegetation on the 
hills is succulent desert, the dry wash in the notch supports riparian vegetation, and the vegeta- 
tion on the surrounding bajada is desert scrub. Species include Larrea tridentata, Chilopsis 
linearis, Fallugia paradoxa, Prosopis glandulosa, Acacia neovernicosa, Brickellia laciniata, 
Yucca torreyi, Agave lecheguilla, Parthenium incanum, and Viguiera stenoloba. 

24) Pine Springs Canyon (el. 1768 m). — Pine Springs Canyon cuts deeply into the eastern 
escarpment of the mountains. The vegetation on the canyon floor is open canyon woodland 
with succulent desert on the slopes. Species include Arbutus xalapensis, Juniperus deppeana, 
Quercus grisea, Bouteloua curtipendula, B. gracilis, and Dasylirion leiophyllum. 

25) Smith Spring-Smith Canyon (el. 1829 m). — Smith Spring is approximately one-third of 
the way up Smith Canyon at the end of a trail starting at Frijole Ranger Station. The vegetation 
around the spring is riparian woodland becoming more open toward the canyon mouth. Plants 
include Juniperus deppeana, Quercus grisea, Bothriochloa sp., Bouteloua gracilis, Lycurus 
phleoides, and Panicum obtusum. 

26) The Bowl (el. 2377 m). — The Bowl is relict coniferous forest interspersed with hard- 
woods. In many places the young trees are growing in very dense almost impenetrable stands. 
The Bowl contains a man-made earthen tank which periodically holds water. Plants in The 
Bowl include Pinus ponderosa, P. strobiformis, Pseudotsuga menziesii, Quercus gambelii, 
Muhlenbergia emersleyi, M. pauciflora, and Agropyron smithii. 

27) Upper Bear Canyon Trail (el. 2362 m).— Upper Bear Canyon Trail is a series of switch- 
backs which traverses a steep rocky slope with thin loose soil. The vegetation is chaparral 
including Quercus undulata, Cercocarpus montanus, and Muhlenbergia pauciflora. 

28) Upper Dog Ranger Station (el. 1920 m).— The ranger station is located on the floor of 
Upper Dog Canyon just north of a point where the canyon becomes considerably narrower. The 



MAMMALS 277 



deep soil of the canyon floor supports open woodland and large, mixed grass meadows. The 
vegetation of the washes on the eastern slope of the canyon and the canyon floor is riparian 
woodland. Open slopes support chaparral and succulent desert vegetation. Riparian woodland 
includes Quercus muhlenbergii, Acer grandidentatum, Arbutus xalapensis, Cercocarpus 
montanus, Quercus undulata, Ceanothus greggii, and Dasylirion leiophyllum. Open slopes 
include Quercus grisea, Cercocarpus montanus, Nolina micrantha, Agave neomexicana, 
Ceanothus greggii, Dasylirion leiophyllum, and Juniperus deppeana. Grasses on the canyon 
floor include Stipa tenuissima, Muhlenbergia repens, Panicum obtusum, and Bouteloua 
gracilis. 

29) Williams Ranch House (el. 1524 m); 30) Vi mi. S, 2 5 / 8 mi. W Guadalupe Peak (el. 1356 
m). — The ranch house is located near the mouth of Bone Canyon at the west base of Guadalupe 
Peak. The vegetation of the bajada around the house is desert scrub, whereas the slopes above 
are characterized by succulent desert vegetation. Succulent desert forms include Agave 
lecheguilla, Dasylirion leiophyllum, Parthenium incanum, and Bouteloua eriopoda. Desert 
scrub includes Larrea tridentata, Prosopis glandulosa, Opuntia lindheimeri, Sporobolus 
contractus, Setaria leucopila, Muhlenbergia porteri, Bouteloua gracilis, and B. eriopoda. 

31) Williams Ranch Road Entrance, 4% mi. S, V% mi. E Guadalupe Peak (el. 1311 m); 32) 4 
mi. S, 1 mi. W Guadalupe Peak (el. 1356 m); 33) 4 mi. S, Vi mi. W Guadalupe Peak (el. 1341 m); 
34) 4 l A mi. S, 1 mi. W Guadalupe Peak (el. 1387 m); 35) 4.3 mi. S Guadalupe Peak (el. 1349 m); 
36) 4 5/ 16 mi. S Guadalupe Peak (el. 1 349 m); 37) 4 3 / 8 mi. S, Vi mi. W Guadalupe Peak (el. 1 356 
m); 38) 4 3 / 8 mi. S, 1% mi. W Guadalupe Peak (el. 1372 m); 39) 4 3 / 8 mi. S, ty% mi. W Guadalupe 
Peak (el. 1372 m); 40) 4Vi mi. S, Vi mi. W Guadalupe Peak (el. 1341 m); 41) 4»/ 2 mi. S, V% mi. E 
Guadalupe Peak (el. 1341 m); 42) 4y 2 mi. S, 3 / 8 mi. E Guadalupe Peak (el. 1326m);43)4 7 / 8 mi.S, 
V% mi. W Guadalupe Peak (el. 1311m); 44) 4 7 / 8 mi. S Guadalupe Peak (el 1326 m); 45) 4 7 / 8 mi. S, 
Vi mi. E Guadalupe Peak (el. 1326 m); 46) 5>/g mi. S Guadalupe Peak (el. 1311m); 47) 5 3 / 8 mi. S, 
Vi mi. W Guadalupe Peak. — Williams Ranch Road enters near the southeastern corner of the 
park. The soil is very sandy in lower areas with patches of higher rockier areas. The area is 
bajada with desert scrub vegetation including Larrea tridentata, Prosopis glandulosa, 
Xanthocephalum sarothrae, Sporobolus contractus, and Bouteloua eriopoda. 

Hudspeth County 

48) Crossroads, 9/ 16 mi. S, 4 5/ 16 mi. W Guadalupe Peak (el. 1219 m); 49) 3 / 8 mi. S, 4 1/16 mi. 
W Guadalupe Peak (el. 1 234 m); 50) 5 / 8 mi. S, 4 3 / 4 mi. W Guadalupe Peak (el. 1 204 m); 5 1 ) 11/ 16 
mi. S, 4 3 / 4 mi. W Guadalupe Peak (el. 1204 m); 52) 7 / 8 mi. S, 4 7 / 8 mi. W Guadalupe Peak (el. 1 196 
m); 53) 7 / 8 mi. S, 4 mi. W Guadalupe Peak (el. 1242 m); 54) 1 mi. S, 4 3 / 4 mi. W Guadalupe Peak 
(el. 1204 m); 55) 1 mi. S, 3 15/16 mi. W Guadalupe Peak.— The Crossroads is the area sur- 
rounding the junction of primitive roads due west of Williams Ranch House and immediately 
north of the central ridge of the Patterson Hills. This area is bajada cut by dry washes. The 
bajada vegetation is desert scrub with riparian vegetation in the washes. Species include Larrea 
tridentata, Prosopis glandulosa, Fallugia paradoxa, Atriplex canescens, Opuntia lindheimeri, 
O. leptocaulis, O. imbricata, and Chilopsis linearis. 

56) Lewis Well, 1 1 / 1 6 mi. S, 6 5 / 8 mi. W Guadalupe Peak (el. 1 128 m); 57) 2 7 / 8 mi. N, 7 7 / 8 mi. W 
Guadalupe Peak (outside of park) (el. 1112 m); 58) V/ 4 mi. W Guadalupe Peak (el. 1 113 m).— 
Lewis Well is an old water well marked by a windmill tower on the bajada near the western 
boundary of the park. West of the well is a large, white, gypsum sand dune, patches of crusted 
gypsum soil, and areas of lacustrine clay. Each of these areas supports distinct plant commu- 
nities. The bajada near the well is dominated by Larrea tridentata and Prosopis glandulosa. 
Lacustrine clay is dominated by Atriplex canescens and crusted gypsum soil is dominated by 
Coldenia hispidissima. The dominant on the gypsum sand dune is Bouteloua breviseta. 

59) Northwest Corner, 4 mi. N, 5'/ 2 mi. W Guadalupe Peak (el. 1 158 m).— The bajada in the 
northwest corner of the park has been grazed far less than any other area of the park west or 
south of the Guadalupe Mountains. This area contains the most well-developed grassland in the 
Hudspeth County portion of the park and may be an example of the potential natural vegeta- 



278 GENOWAYS ET AL. 



tion of the creosote bajada of the park. The vegetation is mixed grassland and succulent desert 
plants including Bouteloua eripoda, Sporobolus cryptandrus, Tridens muticus, Aristidapansa, 
Krameria glandulosa, Erioneuron pulchellum, Dyssodia pentachaeta, Viguiera stenoloba, 
Yucca torreyi, Prosopis glandulosa, Larrea tridentata, Opuntia lindheimeri, O. phaeacantha, 
O. imbricata, O. schottii, O. violacea, O. leptocaulis, and Fouquieria splendens. 

60) Red Sand Dunes, '/ 2 mi. N, 4% mi. W Guadalupe Peak (el. 1 189 m); 61) 1 7/16 mi. N, 5V 2 
mi. W Guadalupe Peak (el. 1 158 m); 62) »/ 2 mi. N, 4»/ 2 mi. W Guadalupe Peak (ei. 1204 m).— 
Near the western boundary of the park, due west of Shumard Peak, is an area of wind-deposited 
quartz sand dunes. The vegetation consists of desert scrub and scattered grasses including 
Prosopis glandulosa, Atriplex canescens, Croton dioicus, Dalea scoparia, D. terminalis, 
Poliomintha incana, Sporobolus contractus, S.flexuosus, S.giganteus, Oryzopsis hymenoides, 
Panicum ramisetum, and Penstemon ambiguus. 

63) Southwest Corner, 1 7 / 8 mi. S, 5 7 / 8 mi. W Guadalupe Peak (el. 1 135 m); 64) V/ 4 mi. S, 5 7 /i mi. 
W Guadalupe Peak (el. 1 151 m); 65) VA mi. S, 5 7/16 mi. W Guadalupe Peak (el. 1 166 m); 66) 
VA mi. S, 5 5/16 mi. W Guadalupe Peak (el. 1 173 m); 67) VA mi. S, 4*4 mi. W Guadalupe Peak 
(el. 1219 m); 68) V/i mi. S, 6V 2 mi. W Guadalupe Peak (el. 1 128 m); 69) 2V4 mi. S, 6 l A mi. W 
Guadalupe Peak (el. 1 1 28 m). — The southwest corner of the westernmost section of park, which 
is due south of Lewis Well, is bajada. This area is subjected to heavy grazing by trespassing cat- 
tle. The vegetation is desert scrub dominated by Larrea tridenta and Atriplex canescens, with 
scattered hummocks of Prosopis glandulosa. 

70) Stage Coach Hills, 9/ 16 mi. S, 4 15/ 16 mi. W Guadalupe Peak (el. 1219 m); 71) V* mi. S, 
5 3 / 8 mi. W Guadalupe Peak (el. 1 1 73 m); 72) Vi mi. S, 5»/ 2 mi. W Guadalupe Peak (el. 1 1 73 m); 73) 
9/16 mi. S, 5 5/16 mi. W Guadalupe Peak (el. 1181 m).— The Stage Coach Hills are a pair of 
small hills north and slightly east of the central ridge of the Patterson Hills. The vegetation of 
the bajada surrounding the hills is desert scrub, whereas succulent desert vegetation is found on 
the hills themselves. Plants include Coldenia hispidissima, C. greggii, Larrea tridentata, Agave 
lecheguilla, Jatropha dioica, Opuntia lindheimeri, O. phaeacantha, Viguiera stenoloba, 
Sporobolus cryptandrus, Bouteloua eriopoda, Muhlenbergia porteri, Hybiscus denudatus, 
Fouquieria splendens, and Selaginella wrightii. 

74) Tank Hill, 1 7/ 16 mi. N, 4'/ 2 mi. W Guadalupe Peak (el. 1234 m); 75) l 3 / 8 mi. N, 4»/ 4 mi. W 
Guadalupe Peak (el. 1227 m). — Tank Hill is an isolated hill north of the Patterson Hills and due 
west of Bartlett Peak. The bajada surrounding the hill was firmly packed quartz and gypsum 
sands with scattered patches of crusted gypsum soil. The vegetation is desert scrub including 
Larrea tridentata, Prosopis glandulosa, Yucca torreyi, Fouquieria splendens, Opuntia 
leptocaulis, Croton dioicus, Melampodium leucanthum, Sporobolus flexuosus S. nealleyi, 
Bouteloua eriopoda, Erioneuron pulchellum, Coldenia hispidissima, and Viguiera stenoloba. 

SPECIES ACCOUNTS 

Myotis californicus californicus (Audubon and Bachman), California Myotis 

Specimens Examined {U).— CULBERSON COUNTY: McKittrick Canyon, 10(TCWC);7mi. 
N Pine Springs, 1 (TCWC); Smith Spring, 2 (TTU). 

The California myotis occurs at intermediate to low elevations within the park, although it 
may not be found in the low bajadas to the west of the mountains. This species has a slow, flut- 
tering flight that can be seen as they forage for insects just at dusk. Daytime retreats sought by 
M. californicus include caves, mines, and rock crevices, where they may also hibernate during 
the colder months of the year. Most of our specimens were taken in mist nets set over water. 

All of the specimens examined are adult males taken in the months of June and August. Nine 
males taken in June had testes that measured 3 in length, whereas two taken in August had testes 
that measured 4. 

We follow Bogan ( 1975) in use of the name combination Myotis californicus californicus for 
bats from this region. Of other species of bats occurring in the park, Myotis californicus is diffi- 
cult to distinguish from Myotis leibii. The braincase arises much more abruptly in M. 



MAMMALS 279 



californicus which is also somewhat paler than M. leibii particularly in coloration of the mem- 
branes. External and cranial measurements of four adult males are as follows: total length, 75, 
76, 74, 83; length of tail vertebrae, 39, 40, 38, 42; length of hind foot, 6, 7, 6, 5; length of ear, 12, 
13, 12, 13; length of forearm, 33.2, 32.0, 3 1.7, 32.8; greatest length of skull, 13.3,13.9, 13.9, 14.1; 
zygomatic breadth, 8.2, 8.4, 8.6, 8.8; postorbital constriction, 3.0, 3.1, 3.1, 3.4; breadth of brain- 
case, 6.2, 6.3, 6.3, 6.6; mastoid breadth, 6.9, 7.2, 7.0, 7.2; length of maxillary toothrow, 4.9, 5.1, 
5.1, 5.1; breadth across upper molars, 5.0, 5.3, 5.2, 5.4. 

Myotis leibii ciliolabrum (Merriam), Small-footed Myotis 

Specimens Examined (5).— CULBERSON COUNTY: Manzanita Spring, 1 (TTU); McKit- 
trick Canyon, 4 (TCWC). 

The small-footed myotis has been obtained only at lower elevations along the eastern escarp- 
ment of the mountains. The species probably can be expected elsewhere in the park at lower ele- 
vations wherever pools of water are available. M. leibii seeks daytime roosts mainly in rock 
crevices and caves and mines. 

The five specimens from the park are adult males which were taken in June. Testes measure- 
ments for four specimens from the early part of the month were all 3, whereas that of the 
specimen taken 23 June was 4. The specimen taken on 2 June at Manzanita Spring was under- 
going annual molt. 

We follow Glass and Baker ( 1968) for use of this trinomial combination. External and cranial 
measurements of the four specimens from McKittrick Canyon are as follows: total length, 78, 
83, 76, 76; length of tail, 38,41,41, 41; length of hind foot, 6, 8, 7, 7; length of ear, 14, 11, 13, 13; 
length of forearm, 30.8,33.0, 32.6, 33.8; greatest length of skull, 13.9, 13.5, 13.6, 14.0; zygomatic 
breadth, 8.5, 8.1, 8.1, 8.7; postorbital constriction, 3.3, 3.1, 3.1, 3.1; breadth of braincase, 6.6, 
6.3, 6.0, 6.6; mastoid breadth, 7.0, 6.6, 6.5, 7.1; length of maxillary toothrow, 5.2, 4.8, 5.1, 5.1; 
breadth across upper molars, 5.5, 5.1, 5.2, 5.3. 

Myotis thysanodes thysanodes Miller, Fringed Myotis 

Specimens Examined (19).— CULBERSON COUNTY: Lost Peak, 2 (TTU); Manzanita 
Spring, 1 (TTU); McKittrick Canyon, 1 1 (TCWC); Smith Spring, 1 (TTU); The Bowl, 4 (TTU). 

The fringed myotis is probably the most common member of the genus occurring in the 
Guadalupe Mountains National Park. The species can be expected to seek daytime shelter in 
rock crevices, man-made buildings, and caves and mines. The majority of our specimens were 
taken in mist nets set over water at night. However, two individuals from Lost Peak were caught 
by hand in an old mine on the opposite side of the ridge from the Lost Peak Mine. This mine is 
about a quarter of a mile deep and is almost horizontal with no side shafts. This species may 
hibernate in caves and mines in the park during the winter months. 

Five adult males taken in late May and June all had testes that measured 3 in length. A male 
taken on 3 July had testes that measured 4, whereas individuals taken on 1 August and 7 August 
had testes that were 3 and 5 in length, respectively. Three adult females captured in early June 
each contained a single embryo that measured as follows in crown-rump length: 1 1 (2 June); 12 
(4 June); 8 (6 June). Two females taken on 6 August were postlactating. A male taken on 7 
August and a female taken on 8 August were nearing completion of annual molt. 

The subspecies thysanodes, originally described from Kern Co., California, is widespread in 
the southwestern United States and clearly includes material from the Guadalupe Mountains 
National Park. External and cranial measurements of two adult males from McKittrick 
Canyon are as follows: total length, 82, 84; length of tail, 4 1 , 40; length of hind foot, 10, 9; length 
of ear, 18, 17; length of forearm, 42.9, 41.4; greatest length of skull, 16.7, 17.0; zygomatic 
breadth, 10.5, 10.5; breadth of postorbital constriction, 4. 1 , 4.2; breadth of braincase, 8.0, 7.8; 
mastoid breadth, 8.2, 8.3; length of maxillary toothrow, 6.0, 6.5; breadth across uppei molars, 
6.6, 6.8. 



280 GENOWAYS ET AL. 



Myotis velifer incautus (J. A. Allen), Cave Myotis 

Specimens Examined (5).— CULBERSON COUNTY: McKittrickCanyon,4(TCWC);4mi. E 
Pine Springs Camp, 1 (TCWC). 

We did not obtain specimens of this species during our studies in the park. However, the 
species can be expected in the park particularly at lower elevations. M. velifer commonly roosts 
in caves and mines; a population is known at Carlsbad Caverns. It is a hibernating species and 
may be expected in the park throughout the year. 

Four of the specimens from the park were adult males taken on the following dates (testes 
measurements in parentheses): 4 June (5); 9 June (6); 31 July (4); 1 August (3). The one adult 
female was not pregnant when taken on 1 August. 

Hay ward (1970) has reviewed geographic variation in Myotis velifer. He concluded, and we 
concur, that populations from west Texas are assignable to the subspecies incautus. External 
and cranial measurements of two adult males are as follows: total length, 88, 96; length of tail, 
40, 45; length of hind foot, 10, 10; length of ear, 15, 14; length of forearm, 43.5, 42.0; greatest 
length of skull, 16.8, 16.2; zygomatic breadth, 10.4, 10.3; postorbital breadth, 3.7, 3.9; breadth 
of braincase, 7.2, 7,3; mastoid breadth, 8.4, 8.2; length of maxillary toothrow, 6.5, 6.3; breadth 
across upper molars, 6.8, 6.7. 

Myotis volans interior Miller, Long-legged Myotis 

Specimens Examined (5).— CULBERSON COUNTY: Manzanita Spring, l(TTU); McKit- 
trick Canyon, 2 (TCWC); The Bowl, 2 (TTU). 

This species of Myotis is evidently not abundant in the Guadalupe Mountains National Park. 
The long-legged myotis occurs in most mountain regions of Trans-Pecos Texas where it 
evidently prefers high, open montane woodlands (Mollhagen and Baker 1972; Davis 1960). 
Easterla (1973a; 1973b) found M. volans to occur in only two plant habitats in the Chisos 
Mountains. These were both woodland habitats found at the highest elevations. Four of our five 
specimens were taken in woodland situations. The fifth specimen was taken at Manzanita 
Spring along the eastern base of the mountains. At this place the vegetation consists of desert 
scrub grassland. 

Our five specimens are all males and were taken on the following dates: 5 June; 10 June; 23 
June; 8 August. Testes measurements for the first two and the last of these individuals were 3, 4, 
and 4. The specimen taken on 23 June was undergoing annual molt. All specimens were netted 
at night over water. 

Myotis v. interior is the trinomial that is applied to populations of this species occurring in the 
west-central United States. Our specimens are definitely included in this taxa. External and 
cranial measurements for two adult males are as follows: total length, 92, 95; length of tail, 44, 
45; length of hind foot, 8, 9; length of ear, 12, 12; length of forearm, 38.8, 38.6; greatest length of 
skull, 14.6, 14.6; zygomatic breadth, 8.9. 8.8; breadth of postorbital construction, 4.0, 3.9; 
breadth of braincase, 7.5, 7.0; mastoid breadth, 7.8, 7.6; length of maxillary toothrow, 5.3, 5.2; 
breadth across upper molars, 5.7, 5.6. 

Lasionycteris noctivagans (Le Conte), Silver-haired Bat 

Specimens Examined ( 18).— CULBERSON COUNTY: McKittrick Canyon, 16 (TCWC); The 
Bowl, 1 (TTU); Thrush Hollow, >/ 4 mi. S Pratt Lodge in South McKittrick Canyon, 1 (TTU). 

This relatively rare species was obtained in relatively high numbers in the Guadalupe Moun- 
tains. The species is known in Texas from only a few localities including the Davis Mountains, 
Bandera County (Davis 1960:51), and on the High Plains. All of these are apparently spring or 
autumn migrants. The population in the Guadalupe Mountains may be in residence during all 
of the summer and, if so, it is the only such population in the state. Our specimens, which are all 
males, were taken during May and June (26 June latest date). 

We have specimens from the montane areas of the park and the riparian woodland areas of 
McKittnjk Canyon. In addition, a specimen was taken just north of the Upper Dog Ranger 
Station at Trail Canyon Tank. The species probably is limited to those areas supporting good 



MAMMALS 28 



stands of trees within the park because silver-hair bats roost in trees. All of our specimens were 
obtained in mist nets set over water where the bats were probably coming to drink. The species is 
migratory and will not be found in the park during the colder months of the year. 

Thirteen adult males captured in the first week of June had testes that averaged 5(4-6) in 
length. None of our specimens evinced molt. External and cranial measurements of five adult 
males from McKittrick Canyon are as follows: total length, 100, 98, 92, 9 1 , 93; length of tail, 45, 
45, 37, 40, 40; length of hind foot, 9, 10, 8, 8, 10; length of ear, 15, 14, 14, 14, 15; length of fore- 
arm, 41.9, 39.6, 38.7, 39.8, 38.8; greatest length of skull, 16.4, 16.1, 16.0, 16.0, 16.3; zygomatic 
breadth, 9.9, 10.0, 9.8, 9.7, 10.2; breadth of postorbital constriction, 4.0, 4.2, 4.1, 4.2, 4.2; 
breadth of braincase, 7.4, 7.8, 7.6, 7.6, 7.9; mastoid breadth, 8.5, 8.6, 8.5, 8.1, 8.8; length of 
maxillary toothrow, 5.7, 5.6, 5.7, 5.4, 5.6; breadth across upper molars, 6.5,6.6,6.6,6.4,6.8. 

Pipistrellus hesperus maximus Hatfield, Western Pipistrelle 

Specimens Examined (23).— CULBERSON COUNTY: Manzanita Spring, 1 (TTU); 
McKittrick Canyon, 13 (TCWC); 7 mi. N Pine Springs, 2 (TCWC); Pratt Lodge, McKittrick 
Canyon, 1 (TTU); Smith Spring, 1 (TTU). HUDSPETH COUNTY: Crossroads, 5 (TTU). 

This is one of the most common and widespread species of bats occurring in the park. One can 
expect to see its fluttering flight anywhere in the park just before darkness during the summer 
months. The only place that populations of this species may be restricted within the park are in 
the high montane areas as we did not obtain specimens of this species in The Bowl even with 
extensive netting. The species does hibernate and, therefore, can be expected to be a year-round 
resident of the park. It roosts during the day in cracks and crevices, mines, and caves. Our five 
specimens from Hudspeth County were shot at dusk as they flew over a dry wash. Vegetation in 
the area consisted of creosote bush, mesquite, four-winged salt bush, and apache plume. All of 
the remaining specimens except those from north of Pine Springs (no information available for 
these) were taken in mist nets set over water. 

Seven males taken in early June had testes that were 3 in length as did males taken on 20 
May, 23 June, and 1 1 July. Males with testes measuring 2 in length were taken on 20 May (2) 
and 22 June. A female taken on 8 June contained two embryos that were 10 in crown-rump 
length. An adult female taken on 31 July was postlactating. 

Geographic variation in the western pipistrelle was studied recently by Findley and Traut 
(1970). They recognized only two subspecies, with the name P. h. maximus being applied to 
populations from east of the Continental Divide. We have followed this arrangement. 

Eptesicus fuscus pallidus Young, Big Brown Bat 

Specimens Examined (30).— CULBERSON COUNTY: Burned Cabin, head McKittrick Can- 
yon, 1 (TCWC); Grisham-Hunter Lodge, McKittrick Canyon, 1 (TTU); Jet. North McKittrick 
Canyon and Devil's Den Canyon, 2 (TTU); Manzanita Spring, 2 (TTU); McKittrick Canyon, 9 
(TCWC); Pine Springs, 1 (TCWC); 2 mi. NW Pine Springs, 2 (TTU); Smith Spring, 2 (TTU); 
The Bowl, 9 (6 TTU, 3 TCWC); Thrush Hollow, % mi. S Pratt Lodge in South McKittrick 
Canyon, 1 (TTU). 

This insectivorous species is one of the most common bat species in the park. All of our speci- 
mens were shot as they foraged at dusk along canyons or were taken in mist nets at night along 
flightways. Although all of our specimens were taken from the top or along the eastern slopes of 
the mountains, this species probably can be expected at any locality in the park where there are 
pools of fresh water suitable for drinking. Because of the flight abilities of this bat, it is easily 
capable of foraging over the low bajadas to the west of the mountains before returning to day- 
time roosts in and near the mountains. Big brown bats will seek daytime shelter in abandoned 
buildings, rock crevices, and old mines (Barbour and Davis 1969). 

In addition to the individuals listed as examined, we banded four bats of this species that were 
netted in The Bowl on the nights of 7 and 8 August 1973. Also two big brown bats were taken at 
Trail Canyon Tank just to the north of the park near Upper Dog Canyon Ranger Station (5.6 
mi. S, 0.6 mi. W El Paso Gap, Eddy Co., New Mexico). 



282 GENOWAYS ET AL. 



All six adult females obtained between 1 June and 23 June were pregnant. Each contained a 
single embryo, which measured 15(1 June), 12 and 14(4 June), 10(6 June), 25(10 June), and 23 
(23 June) in crown-rump length. Testicular lengths of adult male Eptesicus obtained during this 
study were 6 ( 1 June), 8(10 June), 6, 7, and 8 (23 June), 9(12 July), 7(13 July), 8 (6 August), and 
4 (7 August). Adult males undergoing annual molt were taken on 23 June (3 individuals) and 26 
June (1). Two flying young-of-the-year were netted on the night of 7 August in The Bowl. 

We follow the arrangement of Engels (1936) in use of the subspecific name E.f. pallidus for 
brown bats from this area. 

Lasiurus cinereus cinereus (Palisot de Beauvois), Hoary Bat 

Specimens Examined (22).— CULBERSON COUNTY: Manzanita Spring, 1 (TTU); 
McKittrick Canyon, 12 (TCWC); 2 mi. NW Pine Springs, 5 (TTU); The Bowl, 4 (TTU). 

The hoary bat is evidently a common inhabitant of the montane and wooded areas of the 
Guadalupe Mountains during the warmer months of the year. The species is migratory and is 
absent from the area during those times of the year when freezes occur. This species roosts in 
trees and, therefore, is most common in wooded areas; however, it is a strong flier and probably 
could forage throughout the park. We do have one record, Manzanita Spring, that does indicate 
that it forages away from wooded areas on occasion. 

Both sexes are evidently resident in the mountains during at least some of the summer 
months. We have adult males taken on the following dates: 2 June; 3 June; 6 June; 1 1 June; 24 
June; 26 June; 1 August; 2 August; 8 August; 4 September. Adult females, however, have been 
taken only on 3 June and 4 September. The adult males (9) taken in June had testes that 
averaged 4.9 (3-6) in length, whereas those taken in August (5) had testes that measured 6.2 
(5-8). The adult female taken on 3 June in McKittrick Canyon carried two embryos that mea- 
sured 20 in crown-rump length. In addition to the specimens listed above, five individuals of this 
species were taken just north of the park at the Trail Canyon Tank, 5.6 mi. S, 0.6 mi. W El 
Paso Gap, Eddy Co., New Mexico. These five specimens (one male and four females) were 
netted as they came to drink from the tank on the night of 3 June. The four females each 
contained two embryos that ranged from 14 to 17 in crown-rump length. The adult male had 
testes that were 4 in length. One individual of this species (male) was banded and released in The 
Bowl on the evening of 7 August 1973. An adult male netted on 26 June evinced annual molt 
over much of the dorsum. 

The subspecies cinereus has a widespread distribution in North America and is currently the 
only one recognized in this geographic area. 

Plecotus townsendii pallescens (Miller), Townsend's Big-eared Bat 

Specimens Examined { 15). —CULBERSON COUNTY: Lost Peak Mine, 1 (TTU); Manzanita 
Spring, 1 (TTU); McKittrick Canyon, 4 (TCWC); 7 mi. N Pine Springs, 1 (TCWC); Stone 
Cabin, near Grisham-Hunter Lodge, 3 (TTU); The Bowl, 4 (TTU); Upper Dog Ranger Station, 
1 (TTU). 
Additional Record.— CULBERSON COUNTY: Upper Sloth Cave (Davis 1940:74). 

Townsend's big-eared bat is not a common species in the Guadalupe Mountains National 
Park, but it may be expected anywhere in the park at middle and upper elevations. This species 
commonly seeKS refuge in mines or caves during the daytime and will hibernate in them during 
the winter. Two of our specimens— Lost Peak Mine and Upper Dog Ranger Station — were 
obtained from a mine and a small test shaft, respectively, as they slept during the day. The three 
specimens from the Stone Cabin were taken during the daytime as they slept hanging from the 
rafters. The remaining specimens for which we have data were netted over water, including one 
that was banded and released in The Bowl on 7 August 1973. 

A female obtained on 6 August and the one banded in The Bowl were lactating. Other females 
obtained on 3 August, 6 August, and 8 August (2) evinced no reproductive activity. Testes 
lengths for males included the following (date of capture in parentheses): 4 (4 April); 5 ( 1 June); 
6(12 June); 5 (23 June). An adult female taken in The Bowl on 8 August was molting over most 
of its dorsum, whereas another female taken on the same night evinced no molt. 



MAMMALS 283 



Handley (1959) revised the genus Plecotus. He assigned all specimens from Trans-Pecos 
Texas to P. townsendiipallescens, although he considered those living outside of the Guadalupe 
Mountains to be intergrades with P. t. australis. We have followed this arrangement. External 
and cranial measurements of an adult male and female are, respectively, as follows: total length, 
90, 102; length of tail, 48, 45; length of hind foot, 10, 6; length of ear, 33, 37; length of forearm, 
41.8, 42.9; greatest length of skull, 16.4, 16.6; zygomatic breadth, 8.8, 8.8; postorbital constric- 
tion, 3.6, 3.7; breadth of braincase, 7.7, 7.9; mastoid breadth, 9.2, 9.2; length of maxillary 
toothrow, 4.8, 5.3; breadth across upper molars, 5.6, 6.0. 

Antrozous pallidus pallidas (Le Conte), Pallid Bat 

Specimens Examined (24).— CULBERSON COUNTY: >/ 2 mi. NNE Grisham-Hunter Lodge, 
South McKittrick Canyon, 1 (TTU); McKittrick Canyon, 10 (TCWC); 2 mi. NW Pine Springs, 
1 (TTU); 4 mi. E Pine Springs Camp, 4500 ft., 2 (TCWC); Pratt Lodge, McKittrick Canyon, 4 
(TTU); Smith Spring, 3 (TTU); The Bowl, 3 (TTU). 

The pallid bat can be expected throughout the National Park. It probably is a year-round 
resident, hibernating in the colder months of the year. However, our specimens were all taken in 
the four months from May to August. All 24 specimens recorded above are males, which 
indicates that the females are probably forming nursery colonies elsewhere. The pallid bat is 
considered a common inhabitant of the Chihuahuan Desert lowland, but as our records from 
The Bowl indicate, it will range to high altitudes. 

Average testes length for males by month were as follows (range in parentheses followed by 
sample size): May, 5.5 (5-6) 6; June, 5.1 (4-6) 9; July, 5.0 (5) 4; August, 7.7 (6-9) 3. Five 
individuals (two from Pratt Lodge and three from The Bowl) were banded during our studies. A 
specimen taken on 23 June was just beginning annual molt. New hair is evident under the old 
over most of the dorsum of four adult males taken on 11 July at Pratt Lodge. A flying young-of- 
the-year was netted on 7 August 1973 in The Bowl. 

Our specimens are assignable to Antrozous pallidus pallidus as are most other populations of 
pallid bats occurring in the Southwest. 

Tadarida brasiliensis mcxicana (Saussure), Brazilian Free-tailed Bat 

Specimens Examined (16).— CULBERSON COUNTY: McKittrick Canyon, 3 (TCWC); 
Smith Spring, 1 (TTU); The Bowl, 12 (TTU). 

The Brazilian free-tailed bat is a powerful flier and can be expected anywhere in the park. 
However, based on our records this species must confine most of its activity to the montane 
areas and the eastern slopes of the mountains. This bat seeks daytime retreats in caves, mines, 
and old buildings; a large colony, which has been declining in recent years, occupies Carlsbad 
Caverns. Although the population in Carlsbad Caverns includes many adult females and their 
young, all of our specimens are adult males. 

Free-tailed bats have a highly developed migratory pattern and will be found in the park only 
in the months of April to October. All of our specimens were taken in the months of June and 
August. In addition to the specimens listed above, 43 males were banded and released in The 
Bowl on the nights of 7 and 8 August 1973. One specimen also was taken on the night of 4 June 
just north of the Upper Dog Ranger Station at a place designated Trail Canyon Tank, 5.6 mi. S, 
0.6 mi. W El Paso Gap, Eddy Co. , New Mexico. Specimens evincing annual molt were taken on 
23 June and 26 June, although a second specimen taken 26 June was not molting. 

Populations of this species in the western United States and most of Mexico have been 
assigned to T. b. mexicana. This arrangement has been questioned by some recent investigators 
(Cockrum 1969). However, the systematic review of this species has not been published. 

Tadarida macrotis (Gray), Big Free-tailed Bat 

Specimens Examined (13).— CULBERSON COUNTY: McKittrick Canyon, 13 (TCWC). 

This species has been taken in large numbers only in the Chisos (Easterla 1973b: 120) and 
Guadalupe mountains in west Texas; other records for the species in Texas are based on single 



284 GENOWAYS ET AL. 



or a few specimens. Our specimens are all adult females taken on two dates — 11 June 1968 and 3 
August 1970 (LaVal 1973). Our two-year survey has failed to produce additional specimens. 
The specimens were netted as they were coming to drink in wooded areas of lower South 
McKittrick Canyon. However, the area described by LaVal (1973) in which the specimens were 
taken was altered significantly by the floods of 1968. Whether or not a resident population of 
this rare species occurs in the Guadalupe Mountains National Park is not known. However, it is 
clear that a few, possibly migrant, individuals do use the park from time to time. 

Of 12 females taken on 1 1 June, eight carried a single embryo each. The embryos ranged from 
22 to 30 in crown-rump length and averaged 25.9. The female taken on 3 August was lactating. 

This species is considered to be monotypic by modern authors (Husson 1962:258-259). 
External and cranial measurements of five females are as follows: total length, 130, 139, 130, 
129, 126; length of tail, 52, 62, 50, 52, 50; length of hind foot, 12, 13, 12, 12, 12; length of ear, 27, 
30, 29, 29, 27; length of forearm 60.8, 61.5, 59.4, 60.8, 61.3; greatest length of skull, 23.7, 23.7, 
23.9, 23.2, 23.2; zygomatic breadth, 12.5, 12.5, 12.7, 12.5, 12.6; postorbital constriction, 4.2, 4.3, 
4.0, 4.2, 4.2; breadth of braincase, 10.0, 10.0, 10.3, 10.3, 10.3; mastoid breadth, 11.4, 11.6,11.7, 
1 1.4, 11 .5; length of maxillary toothrow, 8.6, 8.6, 8.7, 8.8, 8.3; breadth across upper molars, 8.8, 
8.5,8.8,9.1,9.0. 

Sylvilagus audubonii neomexicanus Nelson, Desert Cottontail 

Specimens Examined (5).— CULBERSON COUNTY: mouth of McKittrick Canyon, 1 
(TCWC); Upper Dog Ranger Station, 4 (TTU). 

The desert cottontail is abundant throughout the park at lower elevations wherever there is 
sufficient cover to provide daytime hiding places. Our specimens from Upper Dog Ranger Sta- 
tion at 1 920 m are from the highest elevation at which the species is presently known in the park. 
Although all of the specimens obtained during our studies are from this location, the species was 
observed at numerous other places including Williams Ranch Road Entrance, along Williams 
Ranch Road, and Patterson Hills Notch in Culberson County, and Lewis Well and the Cross- 
roads in Hudspeth County. Davis (1940:82) reported sighting this species at Pine Springs, West 
Dog Canyon, and along the road at the east base of the mountains during his work in the 
Guadalupe Mountains. 

On 27 July 1973, J. E. Comely obtained a large Crotalus atro x near Choza Spring. Examina- 
tion of the stomach contents of this snake revealed two juvenile S. audubonii each measuring 
approximately 140 in total length. An adult male taken on 31 May 1974 at Upper Dog Ranger 
Station had testes that measured 35 in length. 

We agree with Davis and Robertson (1944:271) that desert cottontails from this part of Texas 
are best assigned to the subspecies neomexicanus. This subspecies was described based upon 
material from Fort Sumner, New Mexico, and is currently applied to specimens from much of 
west Texas and eastern New Mexico. 

Sylvilagus floridanus robustus (Bailey), Eastern Cottontail 

Specimen Examined (1).— CULBERSON COUNTY: The Bowl, 1 (TTU). 

The eastern cottontail may be one of the rarest species of mammals currently occurring in the 
Guadalupe Mountains National Park. Specimens have been recorded only from The Bowl 
(Davis 1940; Davis and Robertson 1944; Hall and Kelson 1951; Hall 195 lb) where it evidently is 
confined to dense stands of Douglas fir and ponderosa pine. During our work in the areas, only 
two individuals were seen and this was only for a brief moment as the rabbit quickly dis- 
appeared into dense underbrush. Davis (1940) estimated that the population of this rabbit was 
approximately 50 individuals. The population is certainly no larger today and may be smaller. 
This taxon occurs only in isolated populations in the Chisos, Chinati, Davis, and Guadalupe 
mountains of Texas. There certainly is no interchange between these populations at the current 
time. 

Our one specimen was a juvenile obtained on 8 June 1 974 as was one specimen taken by Davis 
(1940) on 11 June. 



MAMMALS 285 



The taxonomic status of this taxon is currently uncertain. Beginning with Nelson's revision 
(1909) of the genus, this rabbit was considered a distinct species, S. robustus. This taxonomic 
arrangement prevailed until 1951 when Hall and Kelson (1951:56) presented evidence indicat- 
ing that this rabbit was best considered to be a member of the widespread species S.floridanus. 
Davis (1960) has chosen, however, to retain the specific status for this rabbit under the name 
Sylvilagus robustus. We have chosen to follow Hall and Kelson's revision until further evidence 
is available. Clearly, this rabbit is closely related to S.floridanus but further study may prove its 
specific distinctness. 

Lepus californicus texianus Waterhouse, Black-tailed Jackrabbit 

Specimens Examined (5).— CULBERSON COUNTY: mouth McKittrick Canyon, 5000 ft, 1 
(TCWC); Upper Dog Ranger Station, 2 (TTU); Williams Ranch Road Entrance, 1 (TTU). 
HUDSPETH COUNTY: Lewis Well, 1 (TTU). 

The black-tailed jackrabbit is a common inhabitant of the Chihuahuan Desert portions of the 
park, where its distribution is almost identical with the desert cottontail, Sylvilagus audubonii. 
The highest elevation at which this species has been taken or observed within the park is 1920 m 
in Upper Dog Canyon. In addition to the localities from which specimens were obtained, 
individuals of L. californicus were observed at the following places: Northwest Corner; 
Southwest Corner; Patterson Hills Notch; Stage Coach Hills; Crossroads; Williams Ranch 
House; near Marcus Cabin in West Dog Canyon; near the lower end of Bear Canyon Trail; 
Nipple Hill. Black-tailed jackrabbits are herbivores and are known to forage on grasses and low 
brush. 

The two adult females taken on 2 and 3 June 1973 were pregnant and lactating. The specimen 
taken on 2 June contained five embryos — three in the right uterine horn and two in the left. Two 
of the embryos in the right horn were being reabsorbed. The crown-rump length of the normal 
embryos was 1 1. The female taken on 3 June possessed three embryos in the right uterine horn 
and none in the left. These embryos measured 45 in crown-rump length. A subadult male taken 
on 25 July 1973 at the Williams Ranch Road Entrance had testes measuring 17 in length and was 
molting on the posterior portion of the dorsum and onto the flanks. 

We have assigned our specimens to the taxon Lepus californicus texianus on geographic 
grounds. This name is currently applied to jackrabbits occurring in much of west Texas, New 
Mexico, and north-central Mexico (Hall and Kelson 1959:283). 

Eutamias canipes canipes Bailey, Gray-footed Chipmunk 

Specimens Examined (34).— CULBERSON COUNTY: head of Dog Canyon, 7000 ft, 2 

(USNM); Guadalupe Mts., 7000 ft, 3 (USNM); McKittrick Canyon, 5900 ft, 1 (TCWC); The 

Bowl, 22 (19 TCWC, 3 TTU); Upper Dog Ranger Station, 6 (TTU). 

Additional Record.— CULBERSON COUNTY: Guadalupe Mountains, 8000 ft (Davis 

1940:78). 

The gray-footed chipmunk is confined to the higher elevations of the park; the lowest eleva- 
tion at which the species has been taken is 1800 m in McKittrick Canyon, which is a mesic, 
wooded area. The species is evidently most abundant in and near The Bowl and in Upper Dog 
Canyon. In addition to specimens taken during this study, individuals of this species were 
sighted at Bush Mountain, near Mescalero Campground, and in the upper portions of South 
McKittrick Canyon. All capture sites where the gray-footed chipmunk has been seen are in or 
near forested areas. One specimen was obtained as it climbed in a small Douglas fir tree. The 
Guadalupe Mountains National Park is the only area in Texas where this chipmunk occurs. 

A female taken on 6 August 1973 in The Bowl contained four embryos that measured 28 in 
crown-rump length. Two females taken on 3 and 9 June in Upper Dog Canyon evinced no gross 
reproductive activity. Three male gray-footed chipmunks had the following testes length 
(capture dates in parentheses): 18(31 May); 17(31 May); 5 (9 June). Three individuals taken in 
our study were undergoing molt. A male taken on 3 1 May was molting in a large band across the 
dorsum approximately half-way between the head and rump. The other two individuals were 



286 GENOWAYS ET AL. 



molting in only small areas. A female taken on 3 June was molting on the chest and a female 
taken on 6 August was molting in two small areas on the rump. 

Eutamias canipes is currently considered to occur in restricted montane habitats of the 
Guadalupe, Sacramento, White, Capitan, and Gallinas mountains of Texas and New Mexico. 
There is probably little genetic interchange between isolated populations at the present time. 
Fleharty ( 1 960) recognized a subspecies, E. c. sacramentoensis, as occurring in the Sacramentos 
northward, thus restricting E. c. canipes to the Guadalupe Mountains of Texas and adjacent 
New Mexico. We follow Fleharty's arrangement here. 

External and cranial measurements of four specimens (two males, two females) of E. c. 
canipes deposited in the National Museum of Natural History are as follows (holotype given 
last): total length, 210, 235, 220, 230; length of tail, 96, 105, 97, 104; length of hind foot, 32, 35, 
33, 35; greatest length of skull, 33.9, 36.7, 36.0, 36.5; zygomatic breadth, 18.0, 19.2, 19.7, 19.5; 
interorbital breadth, 7.2, 7.2, 7.8, 8.0; postorbital breadth, 11.4, 11.7. 11.9, ii.6; mastoid 
breadth, 16.2, 16.7, 17.0, 17.2; length of nasals, 10.3, 11.9, 11.4, 11.6; length of maxillary 
toothrow, 5.6, 4.7, 5.6, 5.9; length of palatal bridge, 10.9, 12.1, 11.8, 11.8. 

Ammospermophilus interpres (Merriam), Texas Antelope Squirrel 

Specimens Examined (8).— CULBERSON COUNTY: south of Guadalupe Mountains, 1 
(USNM); mouth McKittrick Canyon, 5000 ft, 1 (TCWC); 7 mi. N Pine Springs, 1 (TCWC); 
below Pine Springs, 2 (TTU); Upper Dog Ranger Station, 3 (TTU). 

Additional Records.— CULBERSON COUNTY: Frijole, about 5600 ft (Davis 1940:77); Pine 
Springs Camp, 5300 ft (Davis 1940:77). 

The Texas antelope squirrel is characteristic of the middle to lower elevations of the Guada- 
lupe Mountains National Park. The species has not been taken or seen at elevations higher than 
1920 m at Upper Dog Ranger Station. Evidently, this squirrel is restricted to rocky areas along 
the escarpment of the mountains as pointed out by Findley et al. (1975:114). Our three 
specimens from Upper Dog were trapped near piles of rock. In addition to the places listed 
above, individuals of this species were sighted on the rocky slope above Williams Ranch House 
in an area dominated by sotol, lecheguilla, and ocotillo, around Nipple Hill, Northwest Corner, 
and near the road immediately below Williams Ranch House. 

None of the specimens taken during our study evinced reproductive activity. Two females 
taken on 5 June 1973 were young of the year. An adult male taken on 30 May 1966 was molting 
in a broad band across the nape of the neck and extending onto the head and shoulders. 

The species Ammospermophilus interpres occupies a relatively restricted geographic range in 
Chihuahua and Coahuila, Mexico, Texas, and New Mexico. The species is relatively uncom- 
mon within the park, but extensive areas of its preferred rocky desert habitat are included in the 
park. Unless major environmental changes occur, this species should present no major man- 
agement problems. 

The five species of the genus Ammospermophilus occupy allopatric geographic ranges. The 
distribution of A. interpres is approached by that of A. harrisii and A. leucurus in New Mexico. 
The relationships of those species (Findley et al. 1975) are currently under investigation at the 
University of New Mexico. For the time being, we considered A. interpres to be a distinct, 
monotypic species. External and cranial measurements of two adult females (mouth of 
McKittrick Canyon and south of Guadalupes) are as follows: total length, 228, 220; length of 
tail, 71, 67; length of hind foot, 40, 39; greatest length of skull, 41.5, 41.1; zygomatic breadth, 
24.3, 23.4; interorbital constriction, 10.3, 9.8; postorbital constriction, 14.1, 14.5; mastoid 
breadth, 20.7, 19.6; length of maxillary toothrow, 6.8, 7.2. 

Spermophilus spilosoma marginatus Bailey, Spotted Ground Squirrel 

Specimens Examined (6).— CULBERSON COUNTY: Williams Ranch Road Entrance, 1 
(TTU). HUDSPETH COUNTY: Tank Hill, 1 7/16 mi. N,4'/ 2 mi. W Guadalupe Peak, 1 (TTU); 
Lewis Well, 4 (TTU). 

The specimens of spotted ground squirrel herein reported are the first known from the park 
area, although Davis and Robertson (1944) reported them from elsewhere in Culberson Co. 



MAMMALS 287 



This species evidently is confined to the Chihuahuan Desert areas of the western bajada. In 
addition to the animals obtained during our study, spotted ground squirrels were sighted near 
the Crossroads and in the Southwest Corner of the park. The specimen taken at Williams Ranch 
Road Entrance was trapped under a creosote bush where the ground was rocky. Two speci- 
mens from Lewis Well were trapped in sandy soil west of the well where gypsum sand dunes 
enter the park. The spotted ground squirrel does not appear to be abundant in the park. 

A female taken at Tank Hill on 15 August 1974 was lactating, whereas two females taken at 
Lewis Well on 18 and 19 May 1974 evinced no reproductive activity. A female from the Williams 
Ranch Road Entrance had four placental scars in the left uterine horn when trapped on 26 July 
1973. A male from Lewis Well had testes that were 17 long on 19 May 1974. This male evinced 
molt on the head and shoulders; the remainder of the pelage evidently was old as it was faded 
and harsh in appearance. 

The type locality for the subspecies marginatus is Alpine, Brewster Co., Texas. Clearly, our 
material from Guadalupe Mountains National Park is indistinguishable from spotted ground 
squirrels from this region of Texas. 

Spermophilus variegatus grammurus (Say), Rock Squirrel 

Specimens Examined ( 17). —CULBERSON COUNTY: Guadalupe Mountains, 2 (USNM); 2 
mi. E mouth of McKittrick Canyon, 5000 ft, 1 (TC WC); % mi. up McKittrick Canyon, 5300 ft, 1 
(TCWC); McKittrick Canyon, 5900 ft, 1 (TCWC); 7 mi. N Pine Springs, 2 (TCWC); l'/ 2 mi. S 
Pine Springs, 1 (TCWC); Pratt Lodge, 1 (TCWC); Upper Dog Ranger Station, 7 (TTU); West 
Dog Canyon, 1 (TCWC). 
Additional Record.— CULBERSON COUNTY: Frijole, about 5600 ft (Davis 1940:76). 

Rock squirrels are the most common sciurid occurring in the park and are abundant particu- 
larly in areas of rock outcroppings. The species was abundant in Upper Dog Canyon during our 
studies where they fed on the berries of Juniperus deppeana and acorns and found refuge in the 
numerous rocky areas. Davis (1940:77) recorded seeing an individual of this species ascend the 
vertical flowering stalk of a century plant and feed on the fruit of the plant. Bailey (1905:85-86) 
records this species as feeding on the berries of Juniperus pachyphloea, acorns of the gray oak, 
cactus fruits (Opuntia engelmanni and Cereus stramineus), and walnuts (Juglans rupestris). In 
addition to the localities listed above, individuals of this species were seen near the Burned 
Cabin at the head of McKittrick Canyon, in Shumard Canyon above the Williams Ranch 
House, and near the Williams Ranch House. On 24 June 1973, a rock squirrel was seen drinking 
from the horse tank in the corral at Frijole. Davis ( 1940) reported seeing this species at 7000 ft 
above Pine Springs Canyon and along the north rim of North McKittrick near the state line. 
Based upon our own and earlier records, therefore, this species occurs as low as 1524 m along 
the west face of the mountains and to at least 5000 ft ( 1 524 m) along the east slope and as high as 
7000 ft (2134 m) in suitable habitats. Bailey (1905) concluded that the species occurred between 
4000 (1220 m) and 7000 ft in the Guadalupes. 

A female taken at Upper Dog Ranger Station on 26 June 1974 contained five embryos that 
measured 1 8 in crown-rump length, whereas another female taken on 26 June 1 973 at this place 
carried four embryos that measured 5. Testes of a male taken on 3 1 May 1974 were 16 long. This 
male was evidently undergoing molt on its head region. The remainder of the pelage appeared to 
be extremely worn. A nonpregnant female taken on 27 July 1974 had completed molt on the 
anterior half of its body and was in the process of molting in the remaining areas, being particu- 
larly evident on its rump. 

The name Spermophilus variegatus grammurus is the scientific name applied to most rock 
squirrels occurring in Trans-Pecos Texas and New Mexico. Our specimens lack the black head 
region as do other members of this subspecies (Howell 1938:143). 

Cynomys ludovicianus (Ord), Black-tailed Prairie Dog 

Specimen Examined (I).— CULBERSON COUNTY: near Guadalupe Mountains, 1 (USNM). 

Although this species once occurred in numerous areas in the vicinity of the Guadalupe 

Mountains, evidently only one melanistic individual that was once held in the National Zoo was 



288 GENOWAYS ET AL. 



ever preserved. The species has been extirpated from the area occupied by the park through the 
direct activity of man. Prairie dogs were eradicated by means of poison because they were 
believed to directly compete with cattle for food in the short-grass prairies. 

Bailey (1905:89-90) reported seeing prairie dogs on the main ridge of the mountains in New 
Mexico and into Dog Canyon in Texas. The name of this canyon was derived from the presence 
of this species. Davis (1940:77-78) did not collect any specimens but did see active colonies at 
the base of Nipple Hill, 3 mi. N of Nipple Hill along U.S. Highway 62- 180, and near the entrance 
of Pine Springs Canyon. He also reported seeing a group of old burrows at the mouth of 
McKittrick Canyon. Clearly, a number of widely scattered colonies of this species once existed 
in the Guadalupe Mountains National Park. A recent attempt by Roger Reisch of the National 
Parks Service to re-introduce this species near Nipple Hill was unsuccessful. If a future attempt 
to re-introduce prairie dogs is planned, the most promising location is the site of an abandoned 
town near Pine Springs where remnants of the old mounds are still evident. At the present time a 
corral for visitors' horses occupies this site. A prerequisite for successful establishment of a new 
prairie dog town would be the relocation of this corral. 

Hollister (1916:19-21), in his revision of the genus, pointed out that the distinction between 
the two subspecies — C. /. ludovicianus and C. 1. arizonensis — of black-tailed prairie dog was 
based upon average differences in cranial measurements and color. He admitted that the sub- 
species were weakly defined and that one individual specimen could not be allotted with any 
certainty. Recently, Pizzimenti (1975) has reviewed members of this genus and he has decided, 
based upon his studies, to consider C. ludovicianus as a monotypic species. We have followed 
this latter arrangement. 

Thomomys bottae guadalupensis Goldman, Botta's Pocket Gopher 

Specimens Examined { 17).— CULBERSON COUNTY: Bear Canyon Pump House, 1 (TTU); 
Burned Cabin, head of McKittrick Canyon, 5 (TCWC); Dog Canyon, 6800 ft, 2 (USNM); 
Manzanita Spring, 1 (TTU); McKittrick Canyon, 4 (USNM); Nipple Hill, 1 (TTU); mouth Pine 
Springs Canyon, 1 (TCWC); Upper Bear Canyon Trail, 2 (TTU). 

This species of pocket gopher occurs at moderate to high elevations within the park. 
Although no specimens were obtained from the top of the mountains, the specimens from 
Upper Bear Canyon Trail were taken near the summit and pocket gopher activity, undoubtedly 
of this species, was noted near the summit of Guadalupe Peak, Bush Mountain, Blue Ridge 
Campground, and Lost Peak. T. bottae occurs in shallow, rocky soil often in association with 
Agave lecheguilla. This pocket gopher frequently feeds on the roots of lecheguilla and will kill 
individual plants. 

Thomomys bottae guadalupensis was described originally by Goldman (1936), with the 
holotype from McKittrick Canyon, although we did not find any pocket gophers in this area 
during our survey. The subspecies was distinguished on the basis of pale coloration and details 
of cranial morphology. As currently understood, this taxon is confined to the Guadalupe 
Mountains. It is worthy of note that we did not find this species to be abundant anywhere within 
the park. However, we believe that this subspecies will be in no danger as long as its preferred 
food of lecheguilla remains abundant. 

External and cranial measurements of two adult males (holotype given first) from McKittrick 
Canyon and two adult females from Burned Cabin, respectively, are as follows: total length, 
218, 218, 195, 200; length of tail, 64, 64, 60, 58; length of hind foot, 29, 29, 29, 28.5; greatest 
length of skull, 38.2, 38.0, 34.6, 37.0; zygomatic breadth, 23.7, 23.3, 21.7, 22.3; interorbital 
breadth, 6.7, 6.8, 7.2, 6.7; squamosal breadth, 19.9, 18.9, 15.0, 15.4; length of nasals, 12.4, 13.8, 
11.3, 13.0; palatal length, 23.8, 24.7, 21.9, 23.2; length of maxillary toothrow, 8.1, 8.5, 7.3, 7.8. 

Pappogeomys castanops parviceps Russell, Yellow-faced Pocket Gopher 

Specimens Examined(5).— CULBERSON COUNTY: 7 mi. N Pine Springs, 1 (TCWC); mouth 
Pine Springs Canyon, 1 (TCWC). HUDSPETH COUNTY: P/ 8 mi. N, 4'/ 4 mi. W Guadalupe 
Peak, 2 (TTU); Lewis Well, 1 (TTU). 



MAMMALS 289 

Additional Record.— CULBERSON COUNTY: foot of Pine Canyon (= Pine Springs Canyon), 
Guadalupe Mts., 5740 ft (Russell 1968). 

During our work in the Guadalupe Mountains National Park, we took specimens of the 
yellow-faced pocket gophers at only two localities, both in the western portion of the park. 
Although we extensively searched the area of Pine Springs Canyon where the species had been 
taken previously, we did not find any evidence that the species currently occurs there. The two 
localities where we trapped members of this taxon were areas of firmly packed quartz and 
gypsum sand, with scattered patches of crusted gypsum soil. The species did not appear to occur 
outside of these soil types at these two places. Our specimens were relatively pale in coloration, 
probably corresponding to the light coloration of the soil in which they lived. 

An adult female taken on 17 August contained a single embryo that measured 40 in crown- 
rump length and a female taken on 7 August carried two embryos that measured 10. Two 
females taken on 14 August and 7 October were nonpregnant. 

External and cranial measurements of four adult females (one from each locality listed, in 
order listed) were as follows: total length, 238, 243, 242, 238; length of tail, 66, 62, 75, 62; length 
of hind foot, 33, 33, 33, 32; condylobasal length, 44.0, 44.5, 43.8, 44.7; zygomatic breadth, 26.6, 
27.7, 27.8, 27.9; interorbital constriction, 7.0, 7.0, 6.6, 6.8; mastoid breadth, 25.8, 25.8, 25.8, 
26.3; squamosal breadth, 20.3, 19.4, 19.2, 19.0; length of nasals, 14.6, 15.4, 15.8, 15.3; length of 
maxillary toothrow, 9.3, 8.9, 8.7, 8.8; palatal length, 29.3, 30. 1, 29.4, 30.2. These measurements 
are in agreement with those given by Russell ( 1968:674) for P. c. parviceps; therefore, we assign 
them to that subspecies. 

Pappogeomys castanops perplanus (Nelson and Goldman), Yellow-faced Pocket Gopher 

Specimens Examined. — None. 

Additional Record.— CULBERSON COUNTY: foot of Pine Canyon (= Pine Springs Canyon), 

Guadalupe Mts., 5740 ft (Russell 1968:653). 

In Russell's (1968) recent revision of the genus Pappogeomys, he divided the subspecies of P. 
castanops into two groups — excelsus subspecies-group and subnubilus subspecies-group. The 
two groups were distinguished mainly on the basis of cranial size, especially measurements of 
cranial length, with the excelsus subspecies-group being much the larger. Russell (1968) 
believed that the two subspecies-groups occurred sympatrically in Pine Springs Canyon with no 
evidence of intergradation. Of the two adult females in a series of five specimens in the Academy 
of Natural Sciences in Philadelphia from the foot of Pine Springs Canyon, he assigned one to P. 
c. perplanus in the excelsus subspecies-group and one (plus the three younger specimens) to P. c. 
parviceps in the subnubilis subspecies-group. He assigned one adult female to P. c. perplanus 
because of its large size (condylobasal length, 48. 2) and the other to P. c. parviceps because of its 
much smaller size (condylobasal length, 45.3). 

During our work in the Guadalupe Mountains, we never obtained P. castanops east of the 
mountains although we searched extensively for their mounds and trapped several Thomomys 
bottae in the area. Changing environmental conditions or intraspecific competition may have 
eliminated this species from east of the mountains in the park at least for the present time. 

Perognathus flavus gilvus Osgood, Silky Pocket Mouse 

Specimens Examined (21).— CULBERSON COUNTY: 3 1/16 mi. S, l 3 / 8 mi. W Guadalupe 
Peak, 1 (TTU); 4 3 / 8 mi. S, P/ 8 mi. W Guadalupe Peak, 1 (TTU); Marcus Cabin, West Dog 
Canyon, 1 (TTU); Nipple Hill, 2 (TTU); Patterson Hills Notch, 1 (TTU); Williams Ranch Road 
Entrance, 9 (TTU). HUDSPETH COUNTY: 1 mi. S, 3 15/ 1 6 mi. W Guadalupe Peak, 1 (TTU); 
Lewis Well, 1 (TTU). 

Additional Record.— CULBERSON COUNTY: 7 mi. N Pine Springs, 5300 ft (Davis and 
Robertson 1944:268). 

The silky pocket mouse occurs in grassland and desert habitats in the park. The highest eleva- 
tion at which we obtained a specimen was 1905 m in West Dog Canyon. Most of our specimens 
were taken on the desert bajadas west and south of the mountains. At the Williams Ranch Road 



290 GENOWAYS ET AL. 



Entrance, where our largest sample was obtained, the vegetation is dominated by creosote bush 
and mesquite with grasses of the genera Sporobolus and Bouteloua. All pocket mice are 
basically granivores and divide the seed resources on the basis of size, availability, and species 
preference. 

A female taken on 17 June evinced two placental scars. Females taken on 19 August, 6 
October, and 7 October were nonpregnant. Testes measurements for males were as follows 
(dates of capture in parentheses): 6 (25 July); 5 (26 July); 4 (28 July); 3 (6 October). Specimens 
taken on 17 June and 25 July were undergoing seasonal molt. In both specimens, molt had 
progressed onto the posterior third of the dorsum. 

Until recently (Wilson 1973), two species of silky pocket mouse— P.flavus and P. merriami— 
were recognized in the area of the Guadalupe Mountains National Park. However, Wilson 
(1973) and later Findley et al. ( 1 975) presented data to show that the two species intergrade and 
that populations previously called P. merriami gilvus were intermediate between the two 
species. We have followed this arrangement pending additional data. Wilson (1973) favored 
retention of the subspecific name gilvus for the intermediate populations. 

The silky pocket mouse is the smallest of the four species of Perognathus occurring in the 
park. It can be distinguished easily from the other species by its overall size and soft silky pel- 
age. External and cranial measurements of three males and two females, respectively, from the 
Williams Ranch Road Entrance are as follows: total length, 104, 113, — , — , 99; length of tail, 
50, 54, —,—, 45; length of hind foot, 12, 16, 16, 11, 15; length of ear, 5, 6, 6, 7, 6; greatest length 
of skull, 20.8, 20.8, 21.0, 19.5, 19.6; zygomatic breadth, 10.6, 11.1, 10.9, 10.2, 10.1; interorbital 
breadth, 4.4, 4.3, 4.3, 4.1, 3.9; mastoid breadth, 12.3, 12.0, 12.3, 11.3, 1 1.4; length of maxillary 
toothrow, 2.8, 3.0, 2.9, 2.9, 2.8; interparetal width, 3.9, 3.3, 3.4, 2.7, 3.1; interparietal length, 2.8, 
2.8, 2.8, 2.8, 2.4. 

Perognathus hispidus paradoxus Merriam, Hispid Pocket Mouse 

Specimen Examined (1).— CULBERSON COUNTY: head of Dog Canyon, 6800 ft, 1 
(USNM). 

The only specimen of this species that has been taken in the park was obtained by Vernon 
Bailey in 1901. This specimen was taken near Bailey's camp at approximately 6800 ft near a 
place that we would term Upper Dog Ranger Station. 

Because this species was not taken during Davis' or our survey, we believe that this species has 
been extirpated from the park. We would suggest that this extirpation may have been caused by 
overgrazing or increasing environmental aridity which have altered the grassy habitat of this 
species. 

This is the largest-sized species of pocket mouse that has occurred in the Guadalupe Moun- 
tains National Park. Measurements for the species are given by Glass (1947). 

Perognathus intermedius intermedius Merriam, Rock Pocket Mouse 

Specimens Examined (30).— CULBERSON COUNTY: Nipple Hill, 1 (TTU); Williams Ranch 

House, 20 (TTU); Williams Ranch Road Entrance, 1 (TTU). HUDSPETH COUNTY: 

Crossroads, 1 (TTU); 1 1 / 16 mi. S, 4% mi.W Guadalupe Peak, 1 (TTU); 3 / 8 mi. S, 4 1 / 16 mi. W 

Guadalupe Peak, 2 (TTU); 1 mi. S, 3 15/16 mi. W Guadalupe Peak, 2 (TTU); Northwest 

Corner, 2 (TTU). 

Additional Record.— CULBERSON COUNTY: 7 mi. N Pine Springs (Davis and Robertson 

1944:268). 

The rock pocket mouse occurs in the grassland and desert habitats of the park, although it is 
evidently most abundant on the desert bajadas west of the mountains. The highest elevation that 
we have taken a specimen is 1646 m at Nipple Hill. This species was taken basically in areas 
where creosote bush, mesquite, and sakbush dominate the vegetation. 

Two nonpregnant females were taken on 15 June and 13 July. 

Perognathus intermedius and P. penicillatus are intermediate in size between the smaller P. 
flavus and larger P. hispidus from the park. The two former species can be distinguished easily 
from the latter two on the basis of external and cranial size. However, we have found it to be 



MAMMALS 291 



extremely difficult to distinguish P. intermedius and P. penicillatus especially in the field. To 
identify our material, we have used the characteristics given by Hoffmeister and Lee 
(1967:367-368). These characteristics seem to separate specimens of the two species from the 
park quite easily in the laboratory. 

We have applied the same P. intermedius intermedius to our specimens following Hall and 
Kelson (1959:501). External and cranial measurements of two males and three females from 
Williams Ranch House are, respectively, as follows: total length, 188, 174, 168, 165,— ; length of 
tail, 109, 100,99,96, — ; length of hind foot, 22, 20, 19, 19, 20; length of ear, 7, 8, 7, 8, 8; greatest 
length ofskull, 25.3, 24.8, 23.2, 23.6, 23.4; zygomatic breadth, 12.3, 12.5, 11.8, 11.6, 12.1; inter- 
orbital breadth, 6.3, 6.3, 6.1, 5.8, 6.1; mastoid breadth, 13.3, 13.2, 12.6, 12.5, 13.0; length of 
maxillary toothrow, 3.8, 3.4, 3.5, 3.5, 3.5; interparietal width, 7.7, 7.0, 7.2, 7.3, 7.2; interparietal 
length, 3.7, 3.4,3.1, 3.0, 2.8. 

Perognathus penicillatus eremicus Mearns, Desert Pocket Mouse 

Specimens £xam/>?^ (83).— CULBERSON COUNTY: '/ 2 mi. S, 5'/ 2 mi. W Guadalupe Peak, 

10 (TTU); Vi mi. S, 4y 4 mi. W Guadalupe Peak, 1 (TTU); 3 1/16 mi. S, l 3 / 8 mi. W Guadalupe 
Peak, 1 (TTU); 3 1 / 4 mi. S, 2 3 / 8 mi. W Guadalupe Peak, 5 (TTU); 4 3 / 8 mi. S, V 2 mi. W Guadalupe 
Peak, 12 (TTU); Nipple Hill, 2 (TTU); Williams Ranch Road, 4 7 / 8 mi. S, V% mi. E Guadalupe 
Peak, 7 (TTU); Williams Ranch House, 1 (TTU); Williams Ranch Road Entrance, 15 (TTU). 
HUDSPETH COUNTY: Crossroads, 11 (TTU); 1 7/ 16 mi. N, 4'/ 2 mi. W Guadalupe Peak, 1 
(TTU); 3 / 8 mi. S, 4 1/16 mi. W Guadalupe Peak, 1 (TTU); 1 1/ 16 mi. S, 4 3 / 4 mi. W Guadalupe 
Peak, 1 (TTU); % m i. s, 4 mi. W Guadalupe Peak, 5 (TTU); l'/ 4 mi. S, 5 7 / 8 mi. W Guadalupe 
Peak, 1 (TTU); 1 '/ 4 mi. S, 5 7/ 16 mi. W Guadalupe Peak, 3 (TTU); l 7 / 8 mi. S, 5% mi. W Guada- 
lupe Peak, 2 (TTU); Lewis Well, 2 (TTU); Stagecoach Hills, 2 (TTU). 

The desert pocket mouse can be expected in the grassland and desert habitats of the park. It is 
most abundant west of the mountains, but has been taken at Nipple Hill east of the mountains. 
Nipple Hill is also the highest elevation at which the species was taken in the park. This species 
occupies much the same habitat in the park as P. intermedius and has been taken with it at six 
localities including Nipple Hill, Williams Ranch Road Entrance, Williams Ranch House, V% mi. 
S, 4 1/ 16 mi. W Guadalupe Peak, 1 1/ 16 mi. S,4 3 4 mi. W Guadalupe Peak, and the Crossroads. 
It is of interest to note that P. penicillatus was the most abundant species at Williams Ranch 
Road Entrance, whereas P. intermedius was the more abundant at Williams Ranch House. 
Generally, P. penicillatus was the more abundant of the two species in the park. 

Females carrying minute embryos were taken on 28 May (3 embryos) and 26 July (no num- 
ber given). None of the four females taken between 5 and 7 October evinced reproductive ac- 
tivity. Testes measurements for males of P. penicillatus were as follows (dates of capture in 
parentheses): 6 (2 1 May); 6(13 July); 4 (28 July); 4 (22 August). A female taken on 26 July was 
undergoing seasonal molt on the posterior portion of the dorsum. 

We follow Hoffmeister and Lee (1967) in application of the name Perognathus penicillatus 
eremicus to desert pocket mice from the park. External and cranial measurements for one male 
and three females, respectively, from Williams Ranch Road Entrance are as follows: total 
length, 152, 155, 164, 164; length of tail, 90, 84, 85, 90; length of hind foot, 22, 25, 22, 21; length 
of ear, 8, 8, 8, 8; greatest length of skull, 25.8, 24.7, 25.2, 25.0; zygomatic breadth, 14.1, 13.1, 
13.2, 13.5; interorbital breadth, 6.7, 6.2, 6.5, 6.3; mastoid breadth, 13.6, 12.7, 12.7, 12.8; length 
of maxillary toothrow, 3.7, 3.5, 3.6, 3.7; interparietal width, 7.5, 7.2, 7.4, 7.0; interparietal 
length, 3.6, 4.0, 3.5, 3.4. 

Dipodomys merriami merriami Mearns, Merriam's Kangaroo Rat 

Specimens Examined (436).— CULBERSON COUNTY: >/ 2 mi. S, 2% mi. W Guadalupe Peak, 

1 1 (TTU); 3 1/16 mi. S, l 3 / 8 mi. W Guadalupe Peak, 7 (TTU); 3>/ 4 mi. S, 2 3 / 8 mi. W Guadalupe 
Peak, 10 (TTU); 4>/ 4 mi. S, 1 mi. W Guadalupe Peak, 37 (TTU); 4 5/ 16 mi. S Guadalupe Peak, 
24 (TTU); 4% mi. S, 1 '/ 8 mi. W Guadalupe Peak, 4 (TTU); 4 3 / 8 mi. S, '/ 2 mi. W Guadalupe Peak, 
56 (TTU); 4'/ 2 mi. S, '/ 2 mi. W Guadalupe Peak, 23 (TTU); 4'/ 2 mi. S, '/s mi. E Guadalupe Peak, 
21 (TTU); 4'/ 2 mi. S, 3 / 8 mi. E Guadalupe Peak, 1 (TTU); 4% mi. S Guadalupe Peak, 13 (TTU); 



292 GENOWAYS ET AL. 



4 7 / 8 mi. S, V% mi. E Guadalupe Peak, 28 (TTU); 5>/ 8 mi. S Guadalupe Peak, 1 (TTU); Patterson 
Hills Notch, 8 (TTU); 7 mi. N Pine Springs, 9 (TCWC); Williams Ranch House, 4 (TTU); 
Williams Ranch Road Entrance, 40 (TTU). HUDSPETH COUNTY: Crossroads, 38 (TTU); 4 
mi. N, 5'/ 2 mi. W Guadalupe Peak, 3 (TTU); 1 7/ 16 mi. N,4'/ 2 mi. W Guadalupe Peak, 6 (TTU); 
'/ 2 mi. N, 43/4 mi. W Guadalupe Peak, 8 (TTU); !/ 2 mi. N, 4>/ 2 mi. W Guadalupe Peak, 6 (TTU); >/ 8 
mi. S, 5% mi. W Guadalupe Peak, 8 (TTU); '/ 2 mi. S, 5'/ 2 mi. W Guadalupe Peak, 16 (TTU); 
11/16 mi. S, 43/4 mi. W Guadalupe Peak, 4 (TTU); 1 mi. S, 3 15/ 16 mi. W Guadalupe Peak, 3 
(TTU); 1 % mi. S, 5 7 / 8 mi. W Guadalupe Peak, 2 (TTU); 1 7 / 8 mi. S, 5 7 / 8 mi. W Guadalupe Peak, 9 
(TTU); Lewis Well, 22 (TTU); Stagecoach Hills, 4 (TTU). 

Merriam's kangaroo rat is the most common kangaroo rat in the park. It is distributed widely 
at lower elevations and is able to utilize the hard rocky desert floor as well as the deeper sandy 
areas. Because of their desert adaptation and adaptation for saltation, kangaroo rats are of 
prime interest to park visitors and individuals could be caged easily in an interpretive center. 
Such interpretive displays in conjunction with an educational program in understanding signs 
made by kangaroo rats should prove of value to the park visitors. 

The diets of Ord's and Merriam's kangaroo rats were the subject of an extensive study pub- 
lished in this volume (O'Connell 1 977) and the interested person is referred to her work. In sum- 
mary, she found the diet of D. merriami to consist of seeds, greenery, and insects. Relative to the 
diet of D. ordii, D. merriami eats greater quantities of insects, especially in the winter months. 

Reproductive data for females are discussed below. In a sample of 27 females collected on 
23 and 24 February, none was pregnant; in a sample of 28 females collected on 22 March, 
none was pregnant; in a sample of six females collected on 20 April, none was pregnant. In a 
sample of 11 females collected on 17 to 21 May, 10 were not pregnant and one female 
contained two embryos with a crown-rump length of 3. In a sample of seven females collected 
on 30 June, five were not pregnant and two contained two embryos each with a crown-rump 
length of 4 and 12. A female collected on 26 July was not pregnant. In a sample of 35 females 
collected on 8 to 23 August, 22 were not pregnant and 13 were pregnant with two embryos 
each. Crown-rump length for the embryos of each female were 2, 3, 4, 15, 18, 23, 24, 24, 29, 29, 
32, 33, and 37. In a sample of 14 females collected on 5 to 7 October, none was pregnant. 
From these data it would appear that the normal number of embryos per litter is two and it 
appears unlikely that a single female produces more than one litter per year. In every case 
observed above, a single embryo was found in each horn of the uterus. The onset of breeding 
appears to be toward the end of May and to cease before October. 

Testicular length for males was as follows (mean, range in parentheses, and number): 
February, 10 (4-13) 24; March, 9 (4-12) 37; April, 8(6-1 1) 5; May, 1 1 (9-12) 1 1; June, 10(5-13) 
1 3; July, 9 (4- 1 2) 8; August, 1 1 (6- 1 3) 3 1 ; October, 5 (4-9) 1 2. The above data suggest that males 
have enlarged testes from February to August during which time most males have scrotal testes. 
During October the testes size is reduced and none of the males had scrotal testes. 

Specimens were observed in molt from February to October. 

We have followed the systematic arrangement of Lidicker (1960) in applying the name 
Dipodomys merriami merriami to our specimens from the park. 

Dipodomys ordii ordii Woodhouse, Ord's Kangaroo Rat 

Specimens Examined (75).— CULBERSON COUNTY: 4J/ 8 mi. S, '/ 2 mi. W Guadalupe Peak, 1 
(TTU); 4% mi. S Guadalupe Peak, 1 (TTU); 5'/ 8 mi. S Guadalupe Peak, 2 (TTU); Williams 
Ranch Road Entrance, 23 (TTU). HUDSPETH COUNTY: Crossroads, 2 (TTU); 2% mi. N, 7 7 / 8 
mi. W Guadalupe Peak, 3 (TTU); l 3 / 8 mi. N, 4'/ 4 mi. W Guadalupe Peak, 1 (TTU); '/ 2 mi. N,4# 
mi. W Guadalupe Peak, 20 (TTU); 7 3 / 4 mi. W Guadalupe Peak, 1 (TTU); 9/ 16 mi. S,5 5/ 16 mi. 
W Guadalupe Peak, 1 (TTU); l'/ 2 mi. S, 6'/ 2 mi. W Guadalupe Peak, 1 (TTU); l 7 / 8 mi. S, 7 7 / 8 mi. 
W Guadalupe Peak, 8 (TTU); Lewis Well, 10 (TTU); Stagecoach Hills, 1 (TTU). 

Our records indicate that Dipodomys ordii is associated with the deeper sandy areas on the 
western side of the park and within this localized habitat the species is relatively abundant. All 
specimens were collected between the elevations of 1230 to 1350 m. 



MAMMALS 293 



The diets of Ord's kangaroo rat and Merriam's kangaroo rat are described in detail in this 
volume (O'Connell 1977), and anyone interested in specifics is referred to her work. Briefly, 
this species eats seeds, greenery, and insects, with the major portion of the diet consisting of 
seeds (mainly of grasses). Ord's kangaroo rat is primarily an opportunistic feeder. 

In a sample of five females collected on 23 and 24 February, only one was pregnant carrying 
an embryo measuring 7 in crown-rump length in each horn of the uterus. Two females taken on 
18 May, a female taken on 29 June, and a female taken on 25 July were not pregnant. In a 
sample of six females taken on 10 to 22 August, two were pregnant. Each pregnant female con- 
tained an embryo in each horn of the uterus with the crown-rump length of the embryos being 4 
for those of one female and 35 for the other. In a sample of three females from 7 October, one 
female contained a minute embryo in each horn and the other two were not pregnant. Testicular 
length for nine males collected in February ranged from 9 to 12, for a male collected in March it 
was 12, for three males collected in May it was 8, 13, and 12. Ten males collected in August had 
testicular lengths that ranged from 8 to 14 and the testicular length of a male collected in 
October was 11. 

Specimens were observed in molt during June and August. 

According to the most recent revision of this species by Setzer (1949), Dipodomys ordii ordii 
is the subspecies occurring in the Guadalupe Mountains National Park. D. ordii and D. 
merriami are similar in external size and coloration; however, the two species are distinguished 
easily because D. ordii has five toes on its hind feet, whereas D. merriami has only four. 

Dipodomys spectabilis baileyi Goldman, Banner-tailed Kangaroo Rat 

Specimens Examined (3).— CULBERSON COUNTY: 4 5/16 mi. S Guadalupe Peak, 1 (TTU). 
HUDSPETH COUNTY: 3/ 8 mi. S, 4 1/ 16 mi. W Guadalupe Peak, 1 (TTU); 1 1 / 16 mi. S, 4% 
mi. W Guadalupe Peak, 1 (TTU). 

Specimens of the banner-tailed kangaroo rat were obtained from the park for the first time 
during our study. Our collecting data suggest that this species is limited to the western and 
southwestern boundaries of the park. This is the largest of the kangaroo rats found in the park 
and is distinguished easily from the other two species by the large tuft of white hairs on the distal 
portion of the tail. Banner-tailed kangaroo rats build conspicuous dens that form mounds with 
several large entrances. Such mounds are infrequent where we observed banner-tailed kan- 
garoo rats in the park but if some mounds prove to be accessible to park visitors, they would 
provide a unique ecological feature for observation. 

A female collected on 23 March did not contain embryos. Testicular length of males was 14 
for a specimen obtained on 29 June and 16 for a specimen from 15 August. Individuals col- 
lected on 29 June and 15 August were molting. 

There has not been a recent systematic review of this species and we follow Hall and Kelson 
(1959) in assigning our specimens to D. s. baileyi. 

Reithrodontomys megalotis megalotis (Baird), Western Harvest Mouse 

Specimens Examined (34).— CULBERSON COUNTY: 4% mi. S, >/ 8 mi. E Guadalupe Peak, 1 
(TTU); Marcus Cabin, West Dog Canyon, 1 (TTU); Pine Springs Campground, 1 (TTU); The 
Bowl, 3 (2 TCWC, 1 TTU); Upper Dog Ranger Station, 27 (TTU); Williams Ranch Road, 4>/ 4 
mi. S, 1 mi. W Guadalupe Peak, 1 (TTU). 

The western harvest mouse occurs at moderate to high elevations throughout the park 
wherever grass occurs. Apparently, it is most common at moderate elevations, but has been 
taken during our study and by Davis (1940:79-80) in grassy meadows in The Bowl. Harvest 
mice are basically granivorous and are probably dependent upon the presence of grass for their 
continued existence in the park. Population estimates for this species in the park are given by 
August et al. (1977). 

A female taken on 19 August in Upper Dog Canyon contained five embryos that measured 16 
in crown-rump length, whereas females taken on 26 January, 6 April, and 24 July evinced no 



294 GENOWAYS ET AL. 



reproductive activity. Testes lengths for males were as follows: April, 4, 4; June, 5, 7, 8; July, 10; 
August, 8, 9. A female taken on 18 August was in subadult pelage. None of our 14 skins was 
from individuals undergoing seasonal molt. 

The subspecies R. m. megalotis is widespread in the western United States and northern 
Mexico. This name has been applied to all members of the species from Trans-Pecos Texas. The 
last systematic revision of the group was by Howell (1914). 

Peromyscus boylii rowleyi (J. A. Allen), Brush Mouse 

Specimens Examined (95).— CULBERSON COUNTY: Bush Mountain, 3 (TTU); '/ 2 mi. 
NNE Grisham-Hunter Lodge, South McKittrick Canyon, 1 (TTU); Guadalupe Mountains, 
7800 ft, 1 (USNM); Guadalupe Mountains, Dog Canyon, 6800 ft, 2 (USNM); Guadalupe 
Mountains, head of McKittrick Canyon, 7800 ft, 1 (USNM); Guadalupe Peak Campground, 1 
(TTU); 4y 8 mi. S, '/ 2 mi. W Guadalupe Peak, 1 (TTU); 5'/ 8 mi. S Guadalupe Peak, 1 (TTU); 
Junction North McKittrick Canyon and Devil's Den Canyon, 1 (TTU); Lost Peak, 1 (TTU); 
Marcus Cabin, West Dog Canyon, 1 (TTU); McKittrick Canyon, 8 (TCWC); Nipple Hill, 1 
(TTU); Pine Springs Campground, 1 (TTU); Pratt Lodge, McKittrick Canyon, 2 (TTU); Smith 
Canyon, 4 (TTU); The Bowl, 30 (19 TCWC, 11 TTU); Upper Dog Ranger Station, 23 (TTU); 
Williams Ranch Road Entrance, 1 (TTU). 

The brush mouse can be expected throughout the park, with the possible exception of the 
lowland desert areas of the western portion. The westernmost record that we have in our 
material is from 4 3 / 8 mi. S, '/ 2 mi. W Guadalupe Peak at an elevation of 1 356 m. The brush mouse 
is evidently most abundant at moderate to high elevations in the park because our largest 
samples are from The Bowl and Upper Dog Ranger Station. During the summer of 1973 when 
we trapped intensively in McKittrick Canyon for more than a week, the rodent populations 
were extremely low, with only four Peromyscus boylii and one P. pectoralis being obtained. 
These specimens were taken on rocky hillsides, where sotol and agave predominated under 
low oak trees. This habitat was similar to a number of areas where the species was taken 
elsewhere in the park. In The Bowl, the brush mouse was taken under large stands of conifers, 
but succulents were not present. A number of specimens were trapped in a log cabin in The 
Bowl. In the area of Williams Ranch Road Entrance, no oaks or conifers were present but the 
brushy vegetation was dominated by creosote bush with scattered mesquite. 

Of the seven species of the genus Peromyscus occurring in the park, P. boylii is evidently the 
most common. Pregnant females of P. boylii were taken on the following dates: 22 March, three 
embryos (4 in crown-rump length); 6 April, minute (no number given); 23 June, two embryos 
(23); 6 August, four embryos (4). Nonpregnant females were obtained in May, June, and July. 
Testes measurements for adult males obtained in our study are as follows: 22 March, 12; 31 
May, 4, 12; 1 June, 5; 2 June, 10;3June, 13;4June, 14; 5 June, 1 1;9 June, 10, 12; 13 June, 8; 23 
June, 1 1; 6 August, 14, 15, 15, 15. Four adult females were found to evince molt on various areas 
of the dorsum on the following dates: 12 June; 23 June; 25 June; 6 August. 

We follow Schmidly (1973) in assigning our specimens to P. b. rowleyi. Relationships of all 
Peromyscus occurring in the park will be discussed in a subsequent publication. 

Peromyscus diffkilis nasutus (J. A. Allen), Rock Mouse 

Specimens Examined (39).— CULBERSON COUNTY: Blue Ridge, 1 (TTU); Blue Ridge 
Campground, 3 (TTU); Bush Mountain, 4 (TTU); Guadalupe Peak Campground, 4 (TTU); 
Guadalupe Mountains, McKittrick Canyon, 7800 ft, 1 (USNM); Lost Peak, 3 (TTU); 
Mescalero Campground, 2 (TTU); Pine Springs Campground, I (TTU); The Bowl, 6 (3 TCWC, 
3 TTU); Upper Dog Ranger Station, 16 (TTU). 

Additional Record.— CULBERSON COUNTY: 2 mi. E Pine Springs (Diersing and 
Hoffmeister 1974:213). 

The rock mouse occurs at moderate to high elevation within the park. Our lowest record of 
occurrence of this species is 1768 m at Pine Springs Campground. However, Diersing and 
Hoffmeister (1974) reported two specimens from 2 mi. E of Pine Springs which is at an approxi- 



MAMMALS 295 



mate elevation of 1 600 m. As the common name of this mouse suggests, it was generally taken in 
rocky situations; at many of the higher elevation localities it was taken sympatrically with P. 
boylii. 

Two females containing embryos were taken during our study. One taken on 3 June carried 
four that measured 4 in crown-rump length and one taken on 1 July carried five that measured 3. 
The average testes length of three males taken on 30 May was 8.3, of six males taken on 9 to 1 1 
June was 9.8, and of a single male trapped on 26 June was 7. Adults that evinced molt were 
taken on 30 May (2), 3 June, and 10 June. 

The rock mouse was first reported as occurring in Texas, based upon specimens from 2 mi. E 
Pine Springs and McKittrick Canyon (Diersing and Hoffmeister 1974:213). Our additional 
specimens confirm the presence of the species and indicate that it is relatively abundant and 
occurs throughout the Guadalupe Mountains of Texas. We follow the systematic arrangement 
of Hoffmeister and de la Torre ( 1961) in using the name Peromyscus difficilis nasutus for these 
mice. Relationships of Peromyscus occurring in the park will be discussed in a subsequent 
publication. 

Peromyscus eremicus eremicus (Baird), Cactus Mouse 

Specimens Examined (42). —CULBERSON COUNTY: Vi mi. S, 2% mi. W Guadalupe Peak, 6 
(TTU); 3 1 / 16 mi. S, 1% mi. W Guadalupe Peak, 1 (TTU); 3'/ 4 mi. S, V/% mi. W Guadalupe Peak, 
1 (TTU); 4 mi. S, '/ 2 mi. W Guadalupe Peak, 2 (TTU); Williams Ranch Road,4'/ 4 mi. S, 1 mi. W 
Guadalupe Peak, 1 1 (TTU); Williams Ranch Road, 4 5/ 16 mi. S Guadalupe Peak, 2 (TTU); 
Williams Ranch Road, 4% mi. S, >/ 8 mi. E Guadalupe Peak, 4 (TTU); 5>/ 8 mi. S Guadalupe Peak, 
1 (TTU); Nipple Hill, 2 (TTU); 7 mi. N Pine Springs, 2 (TCWC). HUDSPETH COUNTY: 
Crossroads, 6 (TTU); % mi. S, 4 1/ 16 mi. W Guadalupe Peak, 3 (TTU); Northwest Corner 1 
(TTU). 

This species occurs in most areas of the xeric lowlands of the park. It is particularly abundant 
in the rocky areas of the western portion of the park where desert scrub vegetation, including 
creosote bush and mesquite, dominates. Many specimens were taken on bajadas where 
Dipodomys merriami was also captured. On the eastern side of the park the vegetation around 
Nipple Hill and north of Pine Springs contained more grasses. 

Three adult females taken on 26 January, 20 May, and 15 August were not pregnant. Three 
adult males had testes lengths of 9, 12, and 11 on 26 January, 15 August, and 23 August, 
respectively. 

We have followed the subspecific arrangement given in Hall and Kelson (1959:607) for this 
species. Relationships of all Peromyscus occurring in the park will be discussed in a subsequent 
publication. 

Peromyscus leucopus tornillo Mearns, White-footed Mouse 

Specimens Examined (20).— CULBERSON COUNTY: '/ 2 mi. S, 2% mi. W Guadalupe Peak, 1 
(TTU); 5'/ 8 mi. S Guadalupe Peak, 2 (TTU); 5 3 / 8 mi. S, '/ 2 mi. W Guadalupe Peak, 1 (TTU); 
Marcus Cabin, West Dog Canyon, 1 (TTU); Pine Springs Campground, 5 (TTU); Williams 
Ranch Road Entrance, 2 (TTU); Williams Ranch Road, 4% mi. S, !/ 8 mi. E Guadalupe Peak, 5 
(TTU). HUDSPETH COUNTY: Crossroads, 1 (TTU); Lewis Well, 1 (TTU); Stage Coach 
Hills, 1 (TTU). 
Additional Record.— CULBERSON COUNTY: Frijole, about 5600 ft (Davis 1940:80). 

The white-footed mouse appears to occur at moderate to low elevations throughout the park, 
but nowhere is it abundant. The species evidently is most common in the grassy areas near the 
Pine Springs Campground and in the desert scrub vegetation along Williams Ranch Road. The 
highest elevation at which we recorded this species was 1905 m in West Dog Canyon where the 
vegetation was mixed grassland with riparian vegetation along the washes. 

An adult female taken on 24 July contained a single embryo that measured 7 in crown-rump 
length; another female taken on 6 October was pregnant but the number of embryos was not 
recorded. Nonpregnant adult females were taken on the following dates: 26 January; 22 March; 



296 GENOWAYS ET AL. 

20 May; 7 October. Three adult males collected on 26 January had testes lengths of 12, 13, and 
13. A specimen taken on 26 January evidenced molt in a small area on the head and neck; the 
remainder of the pelage was adult. 

Our P. leucopus are pale in coloration and should be assigned to P. 1. tornillo which was 
originally described from El Paso, El Paso County, Texas (Mearns 1896). Relationships of all 
Peromyscus occurring in the park will be discussed in a subsequent publication. 

Peromyscus maniculatus blandus Osgood, Deer Mouse 

Specimens Examined (63).— CULBERSON COUNTY: '/ 2 mi. S, 2% mi. W Guadalupe Peak, 1 
(TTU); 3«/ 4 mi. S, 2% mi. W Guadalupe Peak, 1 (TTU); 4 mi. S, '/ 2 mi. W Guadalupe Peak, 5 
(TTU); 4.3 mi. S Guadalupe Peak, 1 (TTU); 4y 8 mi. S, '/ 2 mi. W Guadalupe Peak, 1 (TTU); 
Williams Ranch Road, 4'/ 2 mi. S, 1 mi. W Guadalupe Peak, 2 (TTU); Williams Ranch Road, 4 7 / 8 
mi. S, >/ 8 mi. E Guadalupe Peak, 1 1 (TTU); 5 l / 8 mi. S Guadalupe Peak, 4 (TTU); Williams Ranch 
Road, 5% mi. S, '/ 2 mi. W Guadalupe Peak, 5 (TTU); Nipple Hill, 1 (TTU); Patterson Hills 
Notch, 3 (TTU); 7 mi. N Pine Springs, 1 (TCWC); Williams Ranch Road Entrance, 9 (TTU). 
HUDSPETH COUNTY: Crossroads, 9 (TTU); Lewis Well, 1 (TTU). 

The deer mouse occurs in xeric lowland areas in much the same areas as Peromyscus 
eremicus. On the western bajadas, it was captured among creosote bush and mesquite. The area 
around Lewis Well was somewhat more sandy than other areas in the western portion of the 
park, but our specimen was taken in an area of desert scrub vegetation. Two specimens were 
taken along the eastern front of the mountains in areas that are xeric but contain more grasses 
than the western lowlands. 

Adult males were taken with the following testes measurements (dates of capture in 
parentheses): 9 (22 March); 10 (19 May); 8 (20 May); 7 (30 June); 10 (12 July); 14(13 July); 12 
(26 July); 12, 13 (10 August); 11 (22 August); 10, 11 (6 October). An adult female trapped on 13 
July was carrying three embryos that measured 10 in crown-rump length. Two subadults 
captured on 13 July were molting from subadult to adult pelage. 

Of the 19 skins that we have available, 16 are the light gray coloration typical of Peromyscus 
maniculatus blandus. However, the other three specimens (one each from Williams Ranch 
Road Entrance, 4% mi. S, !/ 8 mi. E Guadalupe Peak, and Crossroads) are predominately buffy 
in color. Because the majority of specimens resemble P. m. blandus in color, we have assigned 
our specimens to this subspecies. Relationships of all Peromyscus occurring in the park will be 
discussed in a subsequent publication. 

Peromyscus pectoralis laceianus Bailey, White-ankled Mouse 

Specimens Examined (20).— CULBERSON COUNTY: 7 mi. N Pine Springs, 15 (TCWC); 
Manzanita Spring, 1 (TTU); Nipple Hill, 2 (TTU); 0.3 mi. N, 0.5 mi. E Pratt Cabin, McKittrick 
Canyon, 1 (TTU); Upper Dog Ranger Station, 1 (TTU). 

We found the white-ankled mouse to be relatively uncommon during our survey of the park's 
mammals. Specimens were obtained at only four localities along the eastern and northern 
boundaries of the park. The habitats in which this species was taken include woodlands and 
grasslands at Nipple Hill and Manzanita Spring and riparian woodland in McKittrick Canyon 
and at Upper Dog Ranger Station. It is probably significant that we did not obtain this species 
at higher elevations or on the desert lowlands of the western portion of the park. 

Adult females taken on 26 January and 1 5 July were nonpregnant. Adult males captured on 3 
June, 23 June, and 29 July had testes that measured 11, 11, and 12, respectively. An adult male 
taken on 23 June was molting over much of its dorsum. 

The most recent systematic review of this species was by Schmidly (1972:1 13-138) and we 
have followed his subspecific arrangement. Relationships of all Peromyscus occurring in the 
park will be discussed in a subsequent publication. 

Peromyscus truei truei (Shufeldt), PirTon Mouse 

Specimens Examined (4).— CULBERSON COUNTY: Marcus Cabin, West Dog Canyon, 3 
(TTU); Upper Dog Ranger Station, 1 (TTU). 



MAMMALS 



297 



Our specimens of the pnfon mouse are the first recorded from the state of Texas. The nearest 
previous record was 15 mi. S Weed, Otero Co., New Mexico, in the southern part of the Sacra- 
mento Mountains (Findley et al. 1975). The only two localities of record are at intermediate 
elevations (1905 m and 1920 m) in the extreme northern portion of the park. Throughout its 
geographic range the pinbn mouse is most common in pfnon-juniper woodlands. Our speci- 
mens were taken in riparian woodlands that included the juniper, Juniperus deppeana, and in 
West Dog Canyon the pinbn pine, Pinus edulis. 

One of our specimens is an adult female that was carrying four embryos measuring 5 when 
captured on 24 July. This specimen was also molting on the lower flanks. A male taken on 2 
June had testes measuring 1 1 . 

The most recent systematic review of this species was by Hoffmeister(1951). Relationships of 
all Peromyscus occurring in the park will be discussed in a subsequent publication. 

Onychomys torridus torridus (Coues), Southern Grasshopper Mouse 

Specimens Examined (63).— CULBERSON COUNTY: '/ 2 mi. S, 2% mi. W Guadalupe Peak, 1 
(TTU); 3>/ 4 mi. S, 2 3 / 8 mi. W Guadalupe Peak, 2 (TTU); 4 mi. S, 1 mi. W Guadalupe Peak, 7 
(TTU); 4'/ 4 mi. S, 1 mi. W Guadalupe Peak, 4 (TTU); 4 5/ 1 6 mi. S Guadalupe Peak, 2 (TTU); 4 7 / 8 
mi. S, Vs mi. E Guadalupe Peak, 16 (TTU); 5'/ 8 mi. S Guadalupe Peak, 5 (TTU); 7 mi. N Pine 
Springs, 1 (TCWC); Williams Ranch Road Entrance, 8 (TTU). HUDSPETH COUNTY: 
Crossroads, 7 (TTU); l 3 / 8 mi. N, 4'/ 4 mi. W Guadalupe Peak, 5 (TTU); 2% mi. S, 6'/ 4 mi. W 
Guadalupe Peak, 2 (TTU); Lewis Well, 2 (TTU). 

The southern grasshopper mouse is found at lower elevations in the desert habitat of the park. 
Because it is almost entirely carnivorous, it is unique among the rodents of the park. Its primary 
food source is insects (Horner et al. 1964; Bailey and Sperry 1929), but other food items include 
scorpions, other arthropods, and mammals. Grasshopper mice have a remarkable behavior 
associated with the killing of prey (Bailey and Sperry 1929; Cole and Wolf 1970; Cyr 1972; 
Horner et al. 1964). With mammals which are nearly as large as grasshopper mice, individuals of 
Onychomys attack this prey from behind and bite them through the cranium, which results in 
instant death. Onychomys also has special means of handling arthropods which have protec- 
tive devices such as scorpions, whip scorpions, and beetles of the genera Eleodes and Chlaenius. 
When attacking scorpions and whip scorpions, they bite off the tail before killing the animal. 
With Eledoes and Chlaenius (beetles which emit defensive secretions), Onychomys grabs these 



■ * 

A 


# * 


* i 


| 


0. 

• 


, t. 


torridus 






X Y 


1 If 

B 


• « 


£ % 


m 


m 


* 9 


W ♦ |p |g 


# 


• 




C 


* % 


m m 


m 






mm » • 


m 


i 


-m • 



Fig. 6. C-band karyotype of Onychomys torridus, male, from the Crossroads, 
Guadalupe Mountains National Park, Texas. 



298 GENOWAYS ET AL. 

beetles with its forepaws and jams its abdomen into the sand to avoid contact with the secre- 
tions. Information concerning the unique predatory behavior of this species would probably 
prove of interest to park visitors who wish to learn more about the park's ecosystem. 

Although limited data have been published on the chromosomal variation in O. torridus, 
data from Baker's laboratory suggest several different chromosomal races occur within this 
species. Therefore the karyotype characteristic of individuals from the park is presented in 
Fig. 6. 

Pregnant females have been collected in June, July, and August. Date of collection, number 
of embryos, and crown-rump length are as follows: 29 June, four embryos with crown-rump 
length of 1 6; 25 July, three embryos of 1 8; 26 July, four embryos of 1 5; 1 August, three embryos 
of 24; 20 August, four embryos of 7. Nonpregnant females were collected on 20 August and on 6 
and 7 October. Date of collection and testicular length (in parentheses) of adult males were i0 
August (11 and 16); 22 August (15); 23 August, (10, 18, and 23); 7 October (5). Adults were 
observed molting in July and October. 

The literature relevant to this species is reviewed by McCarty (1975) and is an excellent source 
for reference to the biology of Onychomys torridus. 

Sigmodon hispidus berlandieri Baird, Hispid Cotton Rat 

Specimens Examined (19).— CULBERSON COUNTY: Guadalupe Mountains, 1 (USNM); 
Guadalupe Mountains, Dog Canyon, 6800 ft, 3 (USNM); Marcus Cabin, West Dog Canyon, 1 
(TTU); Pine Springs Campground, 4 (TTU); Smith Canyon, 3 (TTU); Upper Dog Ranger 
Station, 4 (TTU); Williams Ranch House, 1 (TTU); Williams Ranch Road, Vi mi. S, 2 5 / 8 mi. W 
Guadalupe Peak, 1 (TTU). HUDSPETH COUNTY: Lewis Well, 1 (TTU). 

Hispid cotton rats are known from low to moderate elevations within the park, with Upper 
Dog Canyon being the highest locality from which they have been taken. The species is most 
abundant in the grassy areas along the eastern slopes and northern interior canyons of the 
mountains. However, three specimens were taken along the west slope. Two of these from 
Williams Ranch House area were taken from local grassy situations. The specimens from 
Lewis Well was taken in an Atriplex-Larrea scrub area. The presence of S. hispidus on the 
west side of the Guadalupe escarpment may be an indication of more widespread grasslands at 
an earlier time. This species was not reported previously from within the park. 

Four pregnant hispid cotton rats (number of embryos in parentheses) were taken on 23 June 
(4, 5) and 25 July (3, 4). The following testes measurements were recorded from males captured 
in the park: 26 January, 20; 23 June, 24; 25 June, 20; 23 July, 18; 19 August, 17; 7 October, 17. 
Two juvenile specimens (both females) were taken on 26 January at the Pine Springs Camp- 
ground. Two adult specimens (23 June and 19 August) were undergoing seasonal molt when 
captured. 

The subspecies, S. h. berlandieri, is currently regarded as occurring throughout west Texas 
and most of northern Mexico. The latest systematic review of cotton rats from this area was by 
Bailey (1902a). 

Neotoma albigula albigula Hartley, White-throated Woodrat 

Specimens Examined (38).— CULBERSON COUNTY: Guadalupe Mountains, 1 (USNM); 
Guadalupe Mountains, Dog Canyon, 6800 ft, 1 (USNM); Marcus Cabin, West Dog Canyon, 2 
(TTU); Nipple Hill, 1 (TTU); 7 mi. N Pine Springs, 5300 ft, 7 (TCWC); Upper Dog Ranger 
Station, 20 (TTU); Williams Ranch House, 1 (TTU). HUDSPETH COUNTY: Crossroads, 3 
(TTU); 4 mi. N, 5'/ 2 mi. W Guadalupe Peak, 1 (TTU); Tank Hill, 1 (TTU). 
Additional Record.— CULBERSON COUNTY: Frijole, about 5600 ft (Davis 1940:81). 

Three species of woodrats {Neotoma albigula, N. mexicana, and N. micropus) occur within 
the boundary of the park. An extensive study (Comely 1977) of the ecological distribution of 
these species is published in this volume and we will only briefly summarize his results in the fol- 
lowing three accounts. 

Neotoma albigula occurs around the perimeter of the mountains and on the floors of Upper 
Dog and West Dog canyons which penetrate the mountain mass. On the west side of the park 



MAMMALS 299 

the white-throated woodrat is found primarily in or along edges of dry washes extending west- 
ward from the mountains. The white-throated woodrat is the species which has built most of the 
conspicuous woodrat dens within the park. These dens are constructed of any available material 
and would serve as an excellent item of interest for park visitors. Of particular interest should be 
the role that these nests play as a unique ecological situation that benefits many of the park's 
other species of animals. 

Number and size of embryos and date of collection of pregnant females were as follows: 20 
May, 2 embryos with a crown-rump length of 38; 27 June, 1 embryo with a crown-rump length 
of 37; 29 July, 1 embryo with a crown-rump length of 40; 26 August, 5 embryos with a crown- 
rump length of 2. A post-lactating female was collected on 3 June. Adult females containing no 
embryos were collected on 1 5, 25, and 27 June. Testes length for males and dates collected were 
as follows: 31 May, 7; 2 June, 11, 11; 26 June, 15; 29 June, 19; 23 July, 13; 20 August, 7; 26 
August, 7. Adult specimens were molting on 20 and 27 May, 27 June, and 23 July. 

We follow Hall and Kelson (1959) in assigning these to N. albigula albigula. Even though N. 
albigula and N. mexicana form a contact zone in Upper Dog Canyon, a chromosomal analysis 
failed to reveal any indication of hybridization between the two species. Neotoma albigula and 
N. micropus are also in contact in the southwestern part of the park (Comely 1977); however, 
chromosomal analysis failed to reveal any hybridization in the specimens that we obtained. 

Neotoma mexicana mexicana Baird, Mexican Woodrat 

Specimens Examined (32).— CULBERSON COUNTY: Guadalupe Mountains, 7000 ft, 4 
(USNM); 5'/ 8 mi S Guadalupe Peak, 1 (TTU); The Bowl, 6 (5 TTU, 1 TCWC); Upper Dog 
Ranger Station, 15 (TTU). HUDSPETH COUNTY: % mi. S, 4% mi. W Guadalupe Peak, 1 
(TTU); 11/ 16 mi. S, 4y 4 mi. W Guadalupe Peak, 2 (TTU); 7 / 8 mi. S, 4% mi. W Guadalupe 
Peak, 1 (TTU); 7 / 8 mi. S, 4 mi. W Guadalupe Peak, 1 (TTU); l'/ 4 mi. S, 5 7/16 mi. W 
Guadalupe Peak, 1 (TTU); Lewis Well, 1 (TTU). 

Additional Record.— CULBERSON COUNTY: 7 mi. N Pine Springs, 5300 ft (Davis and 
Robertson 1944:270). 

Neotoma mexicana is distributed throughout Guadalupe Mountains National Park at eleva- 
tions above 1 500 m. The Mexican woodrat is saxicolous and builds its nests in inaccessible rock 
crevices where they would not be observed by park visitors. This species does frequent wooden 
houses such as the cabin in The Bowl and perhaps might be observed by park visitors under such 
circumstances. Comely (1977) has detailed the ecological distribution of this species. Mexican 
woodrats in the log cabin in The Bowl were observed eating acorns which they had gathered in 
large quantities. 

Dates of collection of pregnant females and reproductive data are as follows: 4 June, 2 
embryos (crown-rump length, 1 8), 2 (5); 20 August, 2 (40); 26 August, 3 (22). Adult females con- 
taining no embryos were collected on 3 June and 20 August. Testicular lengths for adult males 
were 19 and 20 for two males collected on 4 June, 17 for a 6 August specimen, and 15 for a speci- 
men from 8 August. Molting was observed for adult specimens collected in June and August. 

We have followed Hall and Kelson (1959) for our systematic arrangement of this species. 

Neotoma micropus canescens J. A. Allen, Southern Plains Woodrat 

Specimens Examined (5).— CULBERSON COUNTY: 4 mi. S, l / 2 mi. W Guadalupe Peak, 2 
(TTU); 43/ 8 mi. S, '/ 8 mi. E Guadalupe Peak, 2 (TTU). HUDSPETH COUNTY: Crossroads, 1 
(TTU). 

The southern plains woodrat has a limited distribution within the park and is first recorded 
for the park based on the specimens collected during our survey (see Comely, this volume for a 
detailed analysis of the habitat of the species). N. micropus is restricted to the lower elevations in 
the southwestern quarter of the park and is locally abundant. As is the case with N. albigula, this 
species builds conspicuous dens under prickly pear, cholla, and possibly other large plants, but 
its limited distribution in the park makes the houses of this species less likely to be v ; ewed by 
park visitors. Although N. micropus and N. albigula are in contact in the southern portion of 
the park, there is virtually no sympatry between them. 



300 GENOWAYS ET AL. 



A female collected on 23 August contained three embryos with a crown-rump iength of 30. 
An adult female collected on 9 August contained no embryos. 

We follow Birney (1973) in assigning our specimens to N. m. canescens. 

Microtus mexicanus guadalupensis Bailey, Mexican Vole 

Specimens Examined (82).— CULBERSON COUNTY: Blue Ridge, 1 (TTU); Guadalupe 
Mountains, 10 (USNM); Guadalupe Peak Campground, 4 (TTU); The Bowl, 41 (32 TCWC, 9 
TTU); Upper Dog Ranger Station, 26 (TTU). 

The Mexican vole is a montane species probably occurring no lower in the mountains than 
1920 m at Upper Dog Ranger Station. Although this species is locally abundant, it is restricted 
to open montane meadows. Because this habitat is limited in the park, the status of this vole and 
its habitat will need continued monitoring. The population of Mexican voles in the Guadalupe 
Mountains National Park is isolated from other populations of the species, with the nearest 
population being on the Sacramento Mountains of New Mexico. This species is best con- 
sidered a relict with Rocky Mountain affinities and is therefore one of the unique features of the 
park. Because of the unique and precarious status of this species, an intensive study of its 
biology has been undertaken by Wilhelm (1977). 

One female taken on 8 August contained three embryos that measured 4 in crown-rump 
length. Males were found to have the following testes measurements (dates of capture in 
parentheses): 7, 3 (5 April); 8.5, 10 (6 April); 1 1 (2 June); 4 (4 June); 9 ( 1 July); 1 1 (21 July); 9 (7 
August); 10 (26 August). A nonpregnant adult female taken on 1 1 June evinced a seasonal molt 
over most of its posterior dorsum. Another female taken on 9 August was molting on the dor- 
sum, but it was impossible to determine whether this was a seasonal or maturational molt. 

The taxon Microtus mexicanus guadalupensis was described originally by Bailey (1902b) on 
the basis of specimens from the Guadalupe Mountains. As this taxon is currently understood, 
populations occurring in the Manzano, Capitan, and Sacramento mountains in New Mexico 
are also included in it. The status of all of these populations is being reviewed by D. E. Wilhelm. 
External and cranial measurements of the male holotype (USNM 109, 191), three adult male 
topotypes, and means (extremes in parentheses) for four adult female topotypes are as follows: 
total length, 152, 145, 147, 150, 142.5 (130-150); length of tail vertebrae, 34, 34, 34, 34, 33.8 
(30-36); length of hind foot, 20, 19, 19, 20, 18.8(18-19); greatest length of skull, 27.1, 25.6, 25.8, 
26.1, 25.2 (24.7-25.8); zygomatic breadth, 16.0, 14.6, 15.2, 15.5, 14.8 (14.4-15.1); interorbital 
constriction, 3.4, 3.1, 3.2, 3.3, 3.2 (3.1-3.3); mastoid breadth, 12.4, 12.0, 11.8, 12.1, 11.6 
(11.3-11.8); length of nasals, 7.1, 7.2, 7.2, 7.5, 7.1 (7.0-7.2); length of maxillary toothrow, 7.3, 
6.2, 6.4, 6.3, 6.3 (6.1-6.9); length of palatal bridge, 5.7, 5.8, 5.5, 5.2, 5.6 (5.4-5.7). 

Erethizon dorsatum couesi Mearns, Porcupine 

Specimens Examined (3).— CULBERSON COUNTY: Bone Springs, 1 (TTU); The Bowl, 1 
(TTU); Upper Dog Ranger Station, 1 (TTU). 

Additional Record.— CULBERSON COUNTY: Burned Cabin, head of McKittrick Canyon, 
7500 ft (Davis 1940:82). 

Although we obtained only three specimens of porcupine during our work in the park, it is 
quite common in the area. It can be expected anyplace in the park where there is sufficient 
woody vegetation to meet its dietary needs. For example, we saw individuals at the north end of 
the Patterson Hills and in the Patterson Hills Notch where some riparian vegetation occurs 
along the washes, with creosote bush and mesquite being the dominant shrubs. Individuals also 
were sighted near Frijole and several places in McKittrick Canyon. The individuals sighted at 
Frijole on 15 May were an adult accompanied by a young. 

This is one of the more conspicuous species of mammal occurring in the park and should be 
included in any interpretative program for the park. Evidence of the activity of this species can 
be seen on many of the trees where they have gnawed away the bark. Porcupines will take refuge 
in trees or rock dens (see Davis 1940:82). Because this species is awkward and slow-moving on 
the ground and unable to escape easily when treed, extra precautions must be taken to protect 
them from park visitors. 



MAMMALS 301 



The specimen from Bone Springs consists of a partial skull that was picked up. The other two 
specimens were nonpregnant females taken on 4 and 9 June. 

We follow Hall and Kelson (1959:782) in assigning our specimens to E. d. couesi on 
geographic grounds. 

Canis latrans texensis Bailey, Coyote 

Specimens Examined. — None. 

Our search of the literature and museum collections has revealed no record of specimens of 
the coyote being taken within the boundaries of the park. Davis and Robertson (1944:265) 
report the species from elsewhere in Culberson County. We heard coyotes at night and received 
reports from park personnel of sightings of coyotes within the park during our work in the area, 
but we did not obtain a specimen. Clearly, this species is present within the park, but has escaped 
collection because of its secretive habits. It is probably one of the more abundant and certainly 
one of the most significant predators occurring within the park. 

Coyotes were heard howling at night by our field parties at the following locations: 
McKittrick Canyon Parking Lot; Williams Ranch Road Entrance; Crossroads; Northwest 
Corner; Red Sand Dunes. Dave Cunningham reported to us that there are coyotes in West Dog 
Canyon and at Coyote Peak. John Chapman reported seeing a coyote cross the road with a 
freshly killed rabbit near McKittrick Canyon, and Comely inspected a pup that had been killed 
on the highway near the Williams Ranch Road Entrance on 13 August 1974. 

Coyotes from the Guadalupe Mountains National Park most likely belong to the subspecies 
C. 1. texensis as this is the subspecies to which Davis and Robertson (1944:265) and Jackson 
(1951:279) assigned other specimens from Culberson County. 

Canis lupus monstrabilis Goldman, Gray Wolf 

Specimen Examined (1). — CULBERSON COUNTY: Guadalupe Mountains, summit of 
mountains near New Mexico line, 1 (USNM). 

The only specimen of the gray wolf from the park is a skull which was obtained by Vernon 
Bailey on 24 August 1901. The following is a quotation from his field notes which are on file at 
the National Bird and Mammal Laboratories, Department of the Interior: "These big wolves 
are said to be especially troublesome in the Guadalupe Mountains and to kill much stock, 
mostly calves and cows. One ranchman said they had killed over 40 head of cattle for him in the 
past three years and that he had been unable to kill any of the wolves. The skull sent in shows 
their size to be very large. The color of this one was light gray." This species has been extirpated 
from the park probably as the result of predator control activities. 

This specimen was assigned to the subspecies C. 1. monstrabilis by Goldman (1944:468). 
Cranial measurements for this specimen are as follows (specimen unsexed but undoubtedly a 
male): condylobasal length, 247.0; zygomatic breadth, 141.0; interorbital constriction, 44.1; 
postorbital constriction, 40.5; mastoid breadth, 84.5; length of nasals, 95.3; length of maxillary 
toothrow, 104.7; palatal length, 128.3. 

Urocyon cinereoargenteus scottii Mearns, Gray Fox 

Specimens Examined (4).— CULBERSON COUNTY: Bear Spring, 5700 ft, 1 (TCWC); 
McKittrick Canyon, 1 (TCWC); The Bowl, 8200 ft, 1 (TCWC); Upper Dog Ranger Station, 1 
(TTU). 

The gray fox is evidently one of the more abundant carnivores occurring in the Guadalupe 
Mountains National Park. All specimens examined were obtained in wooded or canyon situa- 
tions. Our specimen was trapped, using sardines for bait, along the road leading to the ranger 
station in Upper Dog Canyon. Davis (1940:76) reported that the individual from Bear Spring 
was shot as it stalked a cottontail. A fox was observed by Comely on 9 June 1974 near Bush 
Mountain, and another was sighted by Baker at Pine Springs Campground in August 1974. 
The specimen from Upper Dog Ranger Station was an adult male that possessed testes 
measuring 17 in length when taken on 31 May. 



302 GENOWAYS ET AL. 



As currently understood, the name Urocyon cinereoargenteus scottii is applied to gray foxes 
from the park. The subspecies has a widespread occurrence throughout the southwestern 
United States and northern Mexico. External and cranial measurements of a specimen from 
The Bowl (female) and one from Bear Spring (male) are as follows: total length, 940, 1080; 
length of tail, 395, 468; length of hind foot, 129, 142; length of ear, 76, 83; greatest length of skull, 
121.1, 131.3; condylobasal length, 114.7, 128.0; zygomatic breadth, 65.2, 66.9; interorbital con- 
striction, 22.4, 26.7; postorbital constriction, 28.0, 27.9; mastoid breadth, 43.4, 46.2; length of 
maxillary toothrow, 46.4, 53.9; length of palate, 56.9, 63.3. 

Ursus americanus amblyceps Baird, Black Bear 

Specimens Examined. — None. 

Additional Record.~The Bowl (Davis 1940:74). 

In 1901, Bailey found bears to be common on the upper slopes of the almost inaccessible 
canyons of the Guadalupe Mountains. In the head of McKittrick Canyon they had worn paths 
to their feeding areas on the oak and juniper ridges and to waterholes in upper portions of the 
canyon. Evidence of feeding activity of bears was present throughout the upper parts of the 
canyon and on the adjacent ridges. Bailey (1905:188) believed that the bears were feeding on 
acorns, juniper berries, and berries of Berberis fremonti in August. 

Davis (1940:74) estimated the black bear population in the Guadalupe Mountains to be not 
greater than 25 individuals in the late 1930s. He also had reports of the species in upper 
McKittrick Canyon as well as Blue Ridge, Frijole, and the rim of the mountains about 5 miles 
SE of Guadalupe Peak. 

During 1973-74, a bear and bear sign were observed in The Bowl and Upper Dog Canyon. 
Roger Reisch estimated that there was only a single bear in the park at this time. Clearly, the 
population of black bear in the Guadalupe Mountains has declined significantly in recent years. 
This is probably due to hunting pressures. However, with complete protection of the areas 
within the park, the black bear population can be expected to increase again with immigrants 
reaching the area from the Sacramento Mountains in New Mexico where there is a significant 
population. 

The subspecies U. a. amblyceps, which is believed to occur throughout west Texas and New 
Mexico, was described based upon material from Grant Co., New Mexico (see Hall and Kelson 
1959:866-867). 

Ursus arctos Linnaeus, Grizzly Bear 

Specimens Examined. — None. 

The only known specimen of the grizzly bear in Texas is from the Davis Mountains taken in 
1 890 (Bailey 1905; Davis 1 974). It is supposed that this bear may have entered the area from New 
Mexico by way of the Guadalupes. Bailey.( 1932:362-363) believed that specimens "indicate a 
probable range for the species along the Guadalupe, Sacramento, White, Capitan, Manzano, 
and possibly the Jemez Mountains . . ." of New Mexico. He stated that "in 1901, while camped 
at the head of Dog Canyon in the Guadalupe Mountains near the New Mexico and Texas 
boundary line, the writer found tracks of very large bears that were evidently of the grizzly 
group, though apparently no grizzlies had been killed there for some time." Bailey received a 
report from the Forest Service of grizzlies in the Guadalupes in 1909. We believe that there is 
sufficient evidence to include the grizzly bear in the historic mammalian fauna of the Guada- 
lupe Mountains, although there were probably never large numbers of the species in the area. 

Because of the large size of grizzlies and the fact that they kill some livestock, they were 
quickly exterminated from most of their former range. They most certainly were gone from the 
Guadalupes early in this century. 

Numerous species and subspecies have been described for grizzly bears. However, modern 
writers agree that there is only one species involved in the complex. The subspecific arrange- 
ment within the species awaits thorough review. 



MAMMALS 303 



Bassariscus astutus flavus Rhoads, Ringtail 

Specimens Examined (3).— CULBERSON COUNTY: Lower McKittrick Canyon, 0.2 mi. N, 
0.4 mi. W Pratt Lodge, 5150 ft, 1 (TTU); The Bowl, 1 (TCWC); Upper Dog Canyon, 1 (TTU). 

The ringtail is probably quite common in the park but it seldom is seen because it frequents 
rocky, inaccessible habitats. The two specimens that we obtained were found dead, but not as 
the result of our activity. The specimen from The Bowl was a skeleton picked up by Davis 
(1940). Davis (1940) reported signs of this species as being abundant in The Bowl and along the 
cliffs of McKittrick Canyon. He found by examination of the feces that insects constituted a 
large part of the diet of this species. Ringtails have been observed on the stone fence around the 
house at Frijole. 

One of our specimens consists of an unsexed skeleton picked up on 24 June 1974. The other 
specimen (Upper Dog Canyon) was prepared as a standard museum skin and skull. This indi- 
vidual is a nonpregnant adult female found on 29 November 1975. 

Standard cranial measurements of the unsexed individual from The Bowl are as follows: 
greatest length of skull, 81.8; condylobasal length, 79.0; zygomatic breadth, 52.9; interorbital 
constriction, 16.4; postorbital constriction, 16.9; mastoid breadth, 36.0; length of maxillary 
toothrow, 30.7; palatal length, 36.7. We assign our specimens to B. a. flavus which occupies a 
wide geographic range in Texas, Oklahoma, New Mexico, Colorado, and northeastern Mexico 
(Hall and Kelson 1959:881). 

Procyon lotor mexicanus Baird, Raccoon 

Specimens Examined (4).— CULBERSON COUNTY: 0.3 mi. N, 0.5 mi. E Pratt Lodge, 
McKittrick Canyon, 1 (TTU); Upper Dog Ranger Station, 3 (TTU). 

The raccoon apparently has not been reported from the Guadalupe Mountains previously. 
However, we found the species to be relatively abundant in the riparian communities in the 
canyons along the eastern slopes and northern interior canyons of the mountains. They are 
already a nuisance at the Pine Springs Campground, where they raid the trash cans. In addition 
to the places listed, raccoons were sighted in West Dog Canyon, in main McKittrick Canyon, 
and Frijole. This species can be expected anywhere in the park where sources of water are 
associated with wooded areas. 

Two specimens from Upper Dog Ranger Station are unsexed, pick-up skulls. The other two 
specimens are nonpregnant, subadult females. 

We have assigned our specimens to Procyon lotor mexicanus based upon distributional data. 
Goldman (1950:54) assigned a specimen from El Paso to this subspecies; Bailey (1905:194) 
allocated a specimen from Pecos to mexicanus. Based upon this evidence, it seems likely the 
subspecies mexicanus inhabits the Guadalupe Mountains, but the final decision must await the 
obtaining of adult specimens from the area. 

Mustela frenata neomexicana (Barber and Cockerell), Long-tailed Weasel 

Specimens Examined. — None. 

No specimens of the long-tailed weasel were obtained during our work. However, Mr. Roger 
E. Reisch sighted a specimen on or about 23 September 1975 at a place 3.2 mi. S, 3.4 mi. W 
Guadalupe Peak on the Hudspeth-Culberson County line. Although this location is just out- 
side of the park boundary, it clearly indicates that long-tailed weasels are living in the area. In 
addition to sighting the animal, Reisch also collected some fecal material at a presumed den. 
The fecal material is composed almost entirely of insect hard parts. In recent years, long-tailed 
weasels also have been observed in the vicinity of Calsbad Caverns and on ranches adjacent to 
the Guadalupe Mountains National Park. Davis and Robertson (1944) reported a specimen 
from near Kent in southern Culberson County. This individual was noted to be eating a 
woodrat {Neotoma albigula) prior to collection. 

We follow Hall (1951 a:333-338) in assigning long-tailed weasels from this region to Mustela 
frenata neomexicana. 



304 GENOWAYS ET AL. 

Taxidea taxus berlandieri Baird, Badger 

Specimens Examined. — None. 

No specimens of badger were taken during our study. However, the diggings of this species 
were sighted at a number of places, especially near the base of El Capitan in the vicinity of 
Guadalupe Spring. Tony Burgess saw a badger on Williams Ranch Road in the summer of 1973. 
Bailey in his 1901 field notes (on file at National Fish and Wildlife Laboratories) stated that "A 
few badger holes found all over the Mts." Davis and Robertson (1944) recorded several 
sightings in southern Culberson County although no specimens were obtained. Long ( 1 972:750) 
reports a specimen from Carlsbad, Eddy County, New Mexico. Clearly the badger has been, 
and remains, a member of the mammalian fauna of the park. 

We follow the taxonomic arrangement given by Long ( 1972) for the North American badger. 

Spilogale gracilis leucoparia Merriam, Spotted Skunk 

Specimens Examined (2).— CULBERSON COUNTY: Pine Springs, 1 (TCWC); Williams 
Ranch House, 1 (TTU). 

Spotted skunks are relatively rare (possibly due to their secretive habits) throughout Trans- 
Pecos Texas. They are inhabitants of rocky and brushy areas and may be expected wherever 
these occur in the park. Our specimen from Williams Ranch House was trapped in a live trap 
baited with sardines. The trap was placed in a wash immediately above the house. The specimen 
is an adult male that possessed testes measuring 20 in length when taken on 16 June. 

Van Gelder (1959) recognized a single species of spotted skunks in the United States under the 
name S. putorius. Specimens from the Guadalupe Mountains clearly fall within the geographic 
range of S. p. leucoparia as he defined it. Subsequently, Mead (1967, 1968a, b) has presented 
convincing evidence that two species of spotted skunks occur in the United States, with the 
names 5. gracilis for the western species and S. putorius for the eastern. We have chosen to 
follow Mead's evidence for use of the specific name gracilis and have followed Van Gelder's 
use of the subspecific name. 

Mephitis mephitis varians Gray, Striped Skunk 

Specimens Examined (5).— CULBERSON COUNTY: 7 mi. N Pine Springs, 3 (TCWC); Upper 
Dog Ranger Station, 1 (TTU); Williams Ranch House, 1 (TTU). 

Although specimens of striped skunks are available only from intermediate elevations in the 
park, the species has been sighted at the Patterson Hill Notch, near Grisham-Hunter Lodge in 
South McKittrick Canyon, and in The Bowl indicating that the striped skunks may be expected 
anywhere in the park. This species feeds primarily on insects and small vertebrates. As pointed 
out by Findley et al. (1975), the striped skunk is highly susceptible to highway mortality; there- 
fore, with increased vehicular traffic in the park, this species may be affected. 

Both of our specimens from the park are subadults. The female taken on 4 June was non- 
pregnant; the male had testes measuring 20 when taken on 15 June. 

Mephitis mephitis varians occurs from Nebraska to northern Mexico. This was the name 
applied by Davis and Robertson (1944:264) to specimens from the area although they stated 
that specimens from 7 mi. N Pine Springs exhibited some characteristics of M. m. estor which 
occurs to the west. External and cranial measurements of two females from 7 mi. N Pine Springs 
are as follows: total length, 7 1 2, 629; length of tail, 350, 308; length of hind foot, 65, 66; greatest 
length of skull, 70.2, 66.4; condylobasal length, 65.0, 63.2; zygomatic breadth, 41.6, 40.8; in- 
terorbital constriction, 19.1, 20.0; postorbital constriction, 17.9, 18.8; mastoid breadth, 35.6, 
34.9; length of maxillary toothrow, 20.8, 21.6; palatal length, 27.0, 25.6. 

Conepatus mesoleucus mearnsi Merriam, Hog-nosed Skunk 

Specimens Examined (8).— CULBERSON COUNTY: Burned Cabin, head of McKittrick 
Canyon, 1 (TCWC); McKittrick Canyon, 4 (TCWC); The Bowl, 3 (2 TCWC, 1 TTU). 

The hog-nosed skunk may be the most abundant of the three species of skunks occurring in 
the Guadalupe Mountains National Park. Hog-nosed skunks may be expected to occur any- 



MAMMALS 305 

where within the park boundaries, although it will be most abundant in areas of high insect 
populations, which it uses as its main source of food. M embers of this species have been seen at 
Frijole and Blue Ridge Campground in addition to the places listed above. Considerable 
digging activity of this species was noted, during our work, along the road leading to Pratt 
Lodge in McKittrick Canyon. 

Our specimen from The Bowl is a subadult male obtained on 8 June 1974. Testes of this 
individual were 13 in length. 

The subspecies C. m. mearnsi is currently considered to occur throughout most of Texas, 
adjacent New Mexico, and northern Mexico. External and cranial measurements of an adult 
male (Burned Cabin) and female (McKittrick Canyon), respectively, are as follows: total length, 
605, — ; length of tail, 238, — ; length of hind foot, 70, 65; length of ear, 27, 23; greatest length of 
skull, 74.2, 67.7; condylobasal length, 69.2, 62.6; zygomatic breadth, 47.4, 41.2; interorbital 
constriction, 23.2, 21.7; postorbital constriction, 18.7, 18.8; mastoid breadth, 39.9, 36.8; length 
of maxillary toothrow, 21.9, 20.5; palatal length, 29.5, 27.2. 

Felis concolor azteca Merriam, Mountain Lion 

Specimens Examined. — None. 

Evidently no specimens of the mountain lions have been preserved from the Texas portion of 
the Guadalupe Mountains, although the species has occurred there in the past and probably still 
occurs in limited numbers. Bailey in his notes (on file at National Fish and Wildlife Labora- 
tories) stated that mountain lions were "Common in the Mts. where the numerous rock cliffs 
and canyons furnish them excellent cover. Fresh tracks seen above and below our camp in the 
head of Dog Canyon. Panthers are said to kill a good deal of stock, mainly colts, and most of the 
ranchmen keep hounds for hunting them and other 'varments.'" Bailey (1932:289) stated that 
"during 1916 the hunters of the Bureau of Biological Survey killed 9 [mountain lions] in the 
Guadalupe Mountain region," of New Mexico. 

Davis (1940:76) noted mountain lions rarely occurred in the Guadalupe Mountains of Texas. 
He did examine the skin of a mountain lion that had been killed several years earlier near 
Burned Cabin in upper McKittrick Canyon. We believe that several mountain lions (probably 
less than five) have been at least part-time residents of the park in recent years. In 1973, a female 
and two yearlings were reportedly killed just north of the park boundary in New Mexico. These 
lions were allegedly killing sheep and then returning to the safety of the park. In the summer of 
1975, an almost identical incident occurred in the same area. This time an adult lion was cap- 
tured, tranquilized, and removed from the area. One of the problems with the Guadalupe 
Mountains National Park is that it is too small to completely contain the normal home range of 
many of the larger, wide-ranging species such as the mountain lion. During the summer of 1973, 
one of us (Comely) saw large cat tracks, probably of this species, in The Bowl. We also received 
two other reliable reports of mountain lion tracks being seen in The Bowl at other times during 
the same summer. 

The major source of food of this species in the Southwest is mule deer (Davis 1974: 1 34). With 
the increasing population of this food item in the park, the mountain lion can be expected to 
continue to include the park within its current distribution as long as there are populations of 
this cat in areas adjoining the park. 

The taxonomic status of mountain lions occurring in the park is somewhat in question. 
Goldman (1946) in his systematic review of the species assigned specimens from central and 
southern Hudspeth County to F. c. stanleyana. The nearest record to the park was a specimen 
from 25 mi. north of Van Horn (stated to be in Hudspeth County by Goldman). In this same 
work Goldman assigned specimens from New Mexico to F. c. azteca, including one from Queen 
in southwestern Eddy County. Davis (1940) assigned the skin that he had examined from 
Burned Cabin to F. c. azteca; however, without stating a reason Davis and Robertson (1944) 
assigned this same specimen to F. c. stanleyana. We tentatively have assigned the mountain 
lions that occur in the park to F. c. azteca because, based upon all reports that we have received, 
they are entering the park from the north in New Mexico and not from the south. However, 
documented specimens are needed before this assignment can be made definite. 



306 GENOWAYS ET AL. 

Felis rufus bailey i (Merriam), Bobcat 

Specimen Examined (1).— CULBERSON COUNTY: The Bowl, 1 (TCWC). 

The only specimen available from the Guadalupe Mountains is a male collected on 22 June 
1939 in The Bowl. However, we had reports and a sighting of bobcats during our work in the 
area. On 1 6 August 1 973, one of us (Comely) saw a bobcat cross the road just inside the park in 
Upper Dog Canyon at 5:00 p.m. Another bobcat wandered into the yard of the ranger in Upper 
Dog Canyon in the summer of 1973. This individual had numerous porcupine quills embedded 
in its face and obviously had not eaten for a long time. It died after all efforts to help it failed. 
The rangers also reported the presence of bobcats in West Dog Canyon. Several bobcats have 
been trapped in recent years just north of the park in New Mexico according to local ranchers. 
Bailey in his notes (on file at National Fish and Wildlife Laboratories) indicated that bobcats 
were common in the mountains and that he saw numerous tracks and a few skins at ranches. 
Davis ( 1940:76) found by examination of scats that bobcats in The Bowl were living in late June 
almost entirely upon small mammals, especially rabbits. 

We follow Anderson (1972:386-387) in use of the generic name Felis for bobcats previously 
known under the name Lynx. The subspecific name F. r. baileyi has been applied to bobcats 
throughout the southwestern United States and northern Mexico. Cranial measurements of our 
specimen are as follows: greatest length of skull, 1 19.5; condylobasal length, 108.3; zygomatic 
breadth, 83.8; interorbital breadth, 23.9; postorbital breadth, 39.6; mastoid breadth, 55.9; 
length of maxillary toothrow, 35.6; palatal length, 45.8. 

Cervus elaphus merriami Nelson, Merriam's Elk 

Specimens Examined. — None. 

There are apparently no verified records of the native elk in Texas. However, Bailey (1905) 
wrote "several old ranchmen have told me, they ranged south to the southern part of the 
Guadalupe Mountains, across the Texas line. I could not get an actual record of one killed in 
Texas, or nearer than 6 or 8 miles north of the line. . . ." Later, Bailey (1932), writing about New 
Mexico, stated "Merriam's elk is now probably extinct; certainly it no longer occurs in New 
Mexico. Forty years ago it was common in the Sacramento, White, and Guadalupe Mountains 
east of the Rio Grande. . . ." According to Murie (1951), Merriam's elk ranged through only a 
few mountain areas of Arizona, New Mexico, and Texas, isolated by surrounding arid lands. 
We believe that it is relatively safe to include the native elk, C. e. merriami, as a member of the 
mammalian fauna of the Guadalupe Mountains. This subspecies probably became extinct prior 
to 1900 in the area. 

We follow McCullough ( 1969) in use of the specific name C. elaphus for North American elk. 
We have followed McCullough (1969) and Findley et al. (1975) in considering Merriam's elk to 
be a subspecies of the more wide-ranging C. elaphus. The exact taxonomic status of this elk will 
never be determined but our arrangement seems most logical to us. Cervus elaphus merriami 
was apparently larger than C. e. nelsoni and C. e. roosevelti, had more massive antlers, and paler 
coloration. 

Cervus elaphus nelsoni Bailey, Rocky Mountain Elk 

Specimen Examined (1).— CULBERSON COUNTY: Upper Dog Ranger Station, 1 (TTU). 

Forty-four Rocky Mountain elk were introduced into the Guadalupe Mountains in 1928 
(Davis and Robertson 1944). They were imported from the northern Rockies by Judge J. C. 
Hunter and associates and released in McKittrick Canyon. In 1934 (Wright and Thompson 
1934) the herd numbered approximately 60. At that time the elk were concentrated on the slopes 
of McKittrick Canyon near the streambed and were destroying the vegetation. In 1938 (Davis 
1940) the size of the herd was reportedly approximately 400. It is very unlikely that the herd 
could have increased that rapidly. We estimate the present elk population to be 150 or less. 

During our work we observed elk throughout the high country, with the notable exception of 
the Guadalupe Peak-El Capitan area. They have been sighted in Upper Dog Canyon, West Dog 
Canyon, Cox Tank, Frijole, Bush Mountains, and The Bowl. In addition to the above areas, 
Davis (1940) observed elk on Blue Ridge and in McKittrick Canyon. 



MAMMALS 307 

The Guadalupe elk herd is probably the southernmost free-ranging population of Rocky 
Mountain elk. Although the ingestion of succulent vegetation provides some water for the elk 
and mule deer in the Guadalupes, they may be under serious stress from lack of water during the 
driest months of the year. 

In the summer of 1975, two young of the year were observed in Pitchfork Canyon behind 
Upper Dog Ranger Station. Although this is proof that the elk are successfully reproducing, the 
status herd is questionable and is currently the subject of intensive study. Although introduced, 
the elk are a valuable component of the fauna of Guadalupe Mountains National Park. Nothing 
is more exciting to the back country hiker than the sudden appearance of a magnificent bull elk, 
which is one reason why the elk is the one mammal that park visitors often ask about. 

Odocoileus hemionus crooki (Mearns), Mule Deer 

Specimens Examined {!).— CULBERSON COUNTY: McKittrick Canyon, 2 (TCWC); Smith 
Spring, 1 (TTU); The Bowl, 1 (TCWC); Upper Dog Ranger Station, 3 (TTU). 

The mule deer is extremely abundant in the park and its numbers can be expected to continue 
to increase with protection. The major natural predators of mule deer — mountain lions and 
wolves — have been greatly reduced in numbers or eliminated from the region. We observed 
mule deer throughout the park area, but the species was observed most often along the slopes 
and on top of the mountains in areas of dense brush and trees. Mule deer are browsers; their 
food habits have been studied extensively in the New Mexico portion of the Guadalupe Moun- 
tains (Anderson et al. 1965, 1970; Snyder 1961; Kittams et al. 1977). 

All of our specimens are skulls that were picked up from individuals that probably died of 
natural causes. Davis (1940:84) reported a specimen carrying a near full-term fetus when taken 
on 27 June. During our work, a fawn was observed watering at the horse corral at the Upper 
Dog Ranger Station on 29 May. 

We follow Cowan (1956:334) in use of the subspecific name crooki for mule deer from the 
park. Mule deer can be distinguished from the white-tailed deer because their antlers fork 
dichotomously, with prongs being about equal in size, whereas in those of the white-tailed deer 
the prongs appear to arise vertically from a main beam. 

Odocoileus virginianus texanus (Mearns), White-tailed Deer 

Specimens Examined. — None. 

We know of no scientific specimens of the white-tailed deer from the Guadalupe Mountains 
National Park. However, Bailey made the following entry in his field notes during his work in 
the area in 1901 (notes on file with National Bird and Mammal Laboratories, Department of the 
Interior): "A few white-tail deer are said to be found along the east side of the Guadalupe Mts. 
but they are rare. A ranchman who had lived in the Mts. for 1 5 years said he had seen but 3 or 4. 
No doubt they straggle across from the edge of the Staked Plains where they are common." 
During our work in the park, mule deer were found to be extremely common but no white-tailed 
deer were sighted. We can find no justification for the following statement by Davis (1974:257), 
at least for areas within the park boundaries: ". . . in the Guadalupe Mountains the white-tail 
occurs almost entirely in the foothills; the mule deer, in the higher mountains." We believe that 
this species was never abundant in the Guadalupe Mountains and the few individuals present 
were probably extirpated by hunting pressures by man. 

According to Kellogg (1956:35), most of the white-tailed deer in Texas are assignable to the 
subspecies O. v. texanus. We have followed this arrangement. 

Antilocapra americana americana (Ord), Pronghorn 

Specimen Examined (1). — New Mexico Guadalupe Mountains, at east base of mountains, 1 
(USNM). 

The only record of a pronghorn from near the park is based upon a skull picked up by Bailey 
in 190 1 . Writing about this species in his notes (on file with National Bird and Mammal Labora- 
tories, Department of the Interior), Bailey stated: "A few antelope remain on the plains along 
the sides of the Guadalupe Mts. and come up on the foothills and in the side valleys. We saw 



308 GENOWAYS ET AL. 

tracks in Dog Canyon just below our camp. A skull with horns was picked up at the east base of 
the Mts." Nelson (1925) estimated the pronghorn populations of Culberson and Hudspeth 
counties to be 75 and 125, respectively, but none was reported from the area of the park. 
Buechner (1950) plotted the distribution of pronghorns in Trans-Pecos Texas but all herds were 
from either to the south or to the west of the park. Buechner (1950) also presents detailed 
ecological and life history data for this species in Trans -Pecos Texas. Evidently, pronghorns, 
which are basically a grassland species, were never abundant in this area. The species was extir- 
pated from the area probably by hunting pressures or by grazing pressures of cattle. 

The specimen is from near the zone of intergradation between A. a. americana and A. a. 
mexicana. We have followed Bailey's (1932) assignment of this specimen to the former sub- 
species. The skull is that of an adult male, with the following measurements: condylobasal 
length, 275.5; palatal length, 160.5; length of maxillary toothrow, 70.0; squamosal breadth, 
80.7; length of nasals, 107.0; length of horn core, 40.6; breadth of horn core, 23.4. 

Bison bison bison (Linnaeus), Bison 

Specimens Examined. — None. 

According to Allen (1877), the bison did occur in Texas west of the Pecos River but by 1840 
they "no longer ranged west of the Pecos River, either in Texas or New Mexico. . . ." Allen 
(1877:526) reports that on a survey of the area led by Pope in 1854, "Mr. J. H. Byrne, in his diary 
of the expedition, reports meeting bois de vache 'for the first time' at Camp No. 10, near Ojo del 
Cuerbo, or Salt Lakes, west of the Guadaloupe Mountains, and in the Valley of the Rio Grande. 
This is the only allusion to buffalo or buffalo 'sign' contained in the narrative. . . ." Findley et al. 
(1975:335) report a specimen from Carlsbad, Eddy County, New Mexico. 

We believe that the bison did occur during historical times in the area now occupied by the 
Guadalupe Mountains National Park, at least at lower elevations where grassland areas were 
found. The numbers may have never been great and they were probably gone from the area by 
the middle of the 19th century. 

Ovis canadensis mexicana Merriam, Mountain Sheep 

Specimens Examined (3).— CULBERSON COUNTY: McKittrick Canyon, Guadalupe Moun- 
tains, 2 (USNM); Guadalupe Mountains, 1 (USNM). 

Bailey (1905:70-75) reports hunting mountain sheep in the Guadalupe Mountains. The 
stomachs of two individuals that he shot contained Cercocarpus parvif alius, Philadelphia 
microphyllus, common wild onion, and a small amount of grass. Bailey's field notes (on file at 
the National Bird and Mammal Laboratories, Department of the Interior), written during his 
work in August 1901, state that: "Mountain sheep are fairly common in the rough part of the 
range south of Dark Canyon, mainly south of the Texas line. We found where they had been in 
the head of Dog Canyon and McKittrick, and Mr. Frank who lives in Gunsight Canyon told me 
that they were common around his place and in Double Canyon. Mr. Frank has lived in these 
Mts. for about 15 years and has probably killed more sheep than anyone else in the range, 
merely shooting them when they came in sight of his ranch when he needed meat, never more 
than 5 at a time, or 2 in warm weather, as he could not use the meat of a greater number. ... He 
thinks the sheep have increased and are more numerous now than 1 5 years ago. He has counted 
30 in a band but usually finds them in small bands of 5 to 10, sometimes all old rams, or all ewes 
and kids, or in mixed bands." 

Davis and Taylor (1939) and Davis (1940) estimated that no more than 25 bighorns were in 
the Guadalupe Mountains in the late 1930s. They had reports of sightings of mountain sheep in 
1 939 from the east rim of the mountains above Frijole, near El Capitan, north rim of McKittrick 
Canyon, and west rim of mountains near Guadalupe Peak. Davis (1940) also saw two mounted 
heads that were from sheep taken in 1909 on Guadalupe Peak. Gross (1960) summarizes records 
of bighorns in the Guadalupes between 1940 and 1960. Only a few scattered reports were 
received during this time. 

Mountain sheep no longer occur in the Guadalupe Mountains. The species undoubtedly was 
eliminated from the area by the activities of man. The exact causes of their extermination are 



MAMMALS 309 

unknown, but were probably one or a combination of hunting pressures, diseases introduced by 
domestic sheep, or grazing competition of domestic livestock. Of the three specimens preserved 
from the Guadalupe Mountains, two are large, adult males with magnificent sets of horns. The 
other skull is that of a much younger unsexed individual. The Guadalupe Mountains are well 
within the geographic range of O. c. mexicana as currently understood (Hall and Kelson 
1959:1031). 

Serious consideration should be given to a reintroduction of this unique species to the park. 
Sighting this magnificent species would be a treat for the backpackers and hikers and the 
addition of this species would help return the park ecosystem to its original condition. Care 
should be taken concerning the possible origin of the stock for reintroduction, with the reintro- 
duced stock being most like those that were extirpated. A careful evaluation of the extent and 
condition of potential mountain sheep habitat must be completed before reintroduction plans 
are undertaken. Also the potential for spread of disease from domestic sheep to the introduced 
population must be evaluated. 

DISCUSSION 

Our field work and survey of the literature indicate that 65 species of mam- 
mals have occurred in the area now occupied by the Guadalupe Mountains 
National Park during historic times. Of the 65 species, 13 species are bats, 
three rabbits, 29 rodents, 14 carnivores, and six artiodactyls. Another nine 
species possibly occur or possibly have occurred in the park including Notio- 
sorex crawfordi, Myotis evotis, Euderma maculatum, Lasiurus borealis, 
Spermophilus mexicanus, Tamiasciurus hudsonicus, Onychomys 
leucogaster, Vulpes macro tis, and Dicotyles tajacu (an individual, probably 
from an introduced population, was reportedly sighted in the park). 

Extirpated Species 

Of the 65 species known from the park, nine are believed to have been 
extirpated from the area. Most, if not all, of these species have disappeared 
as the direct result of human activity. 

Cynomys ludovicianus. — Bailey (1905) reported prairie dogs to be 
abundant along the main ridge of the mountains in Dog Canyon, which 
derived its name from the numerous colonies of this species in the area. 
Davis (1940) reported scattered "towns" along the eastern edge of the moun- 
tains. The species is no longer present in the park although there are colonies 
in the general area. Prairie dogs were removed from the area by direct human 
activity through the use of poisons because they were believed to directly 
compete with cattle for food. 

Perognathus hispidus paradoxus. — Only one specimen of this species has 
been taken from the park area. Hispid pocket mice are primarily grassland 
inhabitants. The species was probably eliminated from the area by altera- 
tion of this habitat either by overgrazing or environmental changes. 

Canis lupus monstrabilis. — According to Bailey gray wolves were espe- 
cially troublesome in the Guadalupe Mountains and were said to kill much 
stock. One specimen was preserved from the park area. This species was 
extirpated from the park as the result of predator control activities. 

Ursus arctos. — The grizzly bear probably has occurred in the Guadalupe 
Mountains in the past. The species has been removed from much of its 



310 GENOWAYS ET AL. 

former geographic range, probably because of its large size and the fact that 
it does kill some livestock. 

Cervus elaphus merriami. — The geographic range of this extinct sub- 
species of elk once included the Guadalupe Mountains. This subspecies 
probably was removed by the increasing aridity of the region and hunting 
pressures. 

Odocoileus virginianus texanus. — White-tailed deer were never abundant 
in the park area and probably were removed by hunting pressure. 

Antilocapra americana americana. — Bailey reported seeing pronghorn 
along the sides of the Guadalupe Mountains and in the foothills. The species 
was probably never abundant in the area. They probably were eliminated by 
hunting pressure and habitat alteration. 

Bison bison bison. — The bison was probably removed from the area of the 
park by hunting or environmental change by the middle of the 19th century. 

Ovis canadensis mexicana. — Mountain sheep were relatively common in 
the mountains during Bailey's survey in 1901. Herds of 30 individuals were 
reported to Bailey by local ranchmen. Davis, in 1940, estimated that no more 
than 25 bighorns remained in the mountains. Mountain sheep undoubtedly 
were eliminated from the area by a combination of hunting pressures of man, 
diseases introduced by domestic sheep, and grazing competition of domestic 
livestock. 

Species Rare in the Park 

Five species are rare in their distribution in the Guadalupe Mountains 
National Park. Four of these five species are confined to the montane 
regions of the park. This is definitely the most fragile habitat in the park. The 
montane habitat essentially represents an island that is in dynamic 
equilibrium with the Chihuahuan Desert and grassland that surrounds it on 
three sides. 

Sylvilagus floridanus robustus. — This is probably the rarest species still 
occurring in the park. It is evidently confined to the Douglas fir and pon- 
derosa pine stands in The Bowl. The taxon is confined to Chisos, Chinati, 
Davis, and Guadalupe mountains of Texas. There is evidently no inter- 
change between these populations at this time; therefore, if the population in 
the Guadalupe Mountains is lost, no natural repopulation would be 
expected. 

Eutamias canipes canipes. — Gray-footed chipmunks are confined to the 
wooded areas of the higher elevations of the park. This taxon is known only 
from the Guadalupe Mountains; therefore, its existence must be protected. 
E. c. sacrament oensis, the only other subspecies of this species, is known 
only from the Sacramento, White, Capitan, and Gallinas mountains of New 
Mexico, indicating the very restricted distribution of the entire species. 

Ammospermophilus inter pres. — The Texas antelope ground squirrel 
occupies a relatively restricted geographic range in northern Mexico, Texas, 
and New Mexico. Within the park this species occurs in the lower grassland 



MAMMALS 311 

and desert areas. The species is relatively rare within the park, but extensive 
areas of its preferred rocky desert habitat are included in the park. How- 
ever, because these areas will be receiving heavy human impact, the status of 
this unique species should be monitored in the future. 

Thomomys bottae guadalupensis. — This taxon of pocket gopher is 
endemic to the Guadalupe Mountains. Nowhere did we find this gopher to 
be abundant. It is distributed primarily in the montane and valiey areas, but 
we did obtain a specimen near Nipple Hill. We believe that this species will be 
in no real danger as long as its preferred food of lecheguilla remains 
abundant. 

Microtus mexicanus guadalupensis. — This subspecies of the Mexican 
vole is restricted to the Guadalupe Mountains of Texas and the Manzano, 
Capitan, and Sacramento mountains in New Mexico. There is no evidence 
for genetic interchange among these populations at the present time. 
Although this species is locally abundant in the Guadalupe Mountains 
National Park, it is restricted to open montane meadows. Because this 
habitat is limited in the park and subject to heavy human usage, the status of 
the Mexican vole and its habitat will need continued monitoring. 

Mammalian Faunal Relationships within the Park 

Within the park, we recognize four mammalian distributional zones (Fig. 
7). These are based upon major vegetational types (Gehlbach 1967, undated; 
Warnock undated) and the distribution of some indicator species of mam- 
mals. For a species to be a good indicator, it should be relatively abundant 
and its distribution should be restricted, or nearly so, to the zone for which it 
is an indicator. The four zones that we recognize and their indicator species 
of mammals are as follows: desert — Dipodomys merriami, D. speciabilis, 
Spermophilus spilosoma, Onychomys torridus, and Neotoma micropus; 
grassland — Perognathus hispidus, Sigmodon hispidus, and Reithro- 
dontomys megalotis; riparian woodland — Procyon lotor; montane — 
Sylvilagus floridanus, Peromyscus boylii, Neotoma mexicana, and 
Microtus mexicanus. 

The grassland habitat was found to contain the most mammalian species 
(41) and the montane the least with 27 (Table 1). The number of mammalian 
species shared between habitats is shown in the upper portion of Table 1 . The 
highest Burt Coefficient of Similarity between habitats was between the 
Montane and Riparian Woodland. The Grassland Zone had a relatively 
high coefficient with all habitats. The lowest coefficient was between the 
Montane and Desert zones which share only eight mammalian species. 
Clustering of the Burt Coefficients (Fig. 8) using the Unweighted Pair Group 
Method Using Arithmetic Means groups Desert and Grassland mammalian 
faunas together and the Riparian Woodland clusters closely with the 
Montane. 

These results indicate that the Desert and Montane mammalian faunas 
are quite distinct. The Grassland mammalian fauna seems to be transitional 



312 



GENOWAYS ET AL. 




Fig. 7. Mammalian distributional zones in the Guadalupe Mountains National 
Park, Texas. See text for discussion. 



TABLE 1. Similarity of mammalian faunas occurring in the four mammalian distributional 
zones recognized in the Guadalupe Mountains National Park, Texas. Boldface numbers on 
the diagonal represent the total number of mammalian species occurring in each zone. The 
numbers above the diagonal represent the number of species shared between zones, whereas 
the numbers below the diagonal represent the Burt Coefficients of Similarity between the 



Distributional 
Zone 



Desert 



Grassland 



Riparian 
woodlands 



Montane 



Desert 


31 


24 


12 


8 


Grassland 


67 


41 


24 


18 


Riparian Woodland 


39 


67 


31 


22 


Montr ne 


28 


53 


76 


27 



MAMMALS 313 



£ 



Desert 
Grassland 

Riparian Woodland 
Montane 



45.9 54.9 68.4 77.4 

Fig. 8. Phenogram resulting from the clustering (Unweighted Pair Group Method 
Using Arithmetic Means) of Burt Coefficients of Similarity among mammalian dis- 
tributional zones as given in Table 1. 



between that of the Desert and Montane-Riparian Woodland faunas. This 
would account for high number of species in the grasslands and high coeffi- 
cients with all other faunas. However, we have chosen to recognize this zone 
because there are some mammalian species which are limited to the grass- 
land and would probably be eliminated from the park if the Grasslands are 
eliminated. The Montane and Riparian Woodland mammalian faunas are 
the most similar. This is reasonable because the habitats are in close geo- 
graphic proximity and both represent relatively mesic habitats. 

It is the Montane mammalian fauna that gives the Guadalupe Mountains 
National Park its unique character. Several species present in this area are 
at, or near, the southern limit of their distribution and represent a southern 
attenuation of the Rocky Mountain fauna of New Mexico. This montane 
island at the edge of the Chihuahuan Desert, and in dynamic equilibrium 
with it, is one reason that park was preserved and is the reason that a sound 
management plan must be developed and followed for the park. 



Comparisons of Mammalian Faunas from Guadalupe Mountains National 
Park with those from Other Geographic Areas 

We compared the mammalian fauna of the Guadalupe Mountains 
National Park with other specific areas from which we believe a relatively 
complete mammalian fauna is known. We chose areas within the same geo- 
graphic region as the Guadalupe Mountains as follows (references in 
parentheses are those used to develop the faunal list given in Table 2): Big 
Bend National Park, Texas (Borell and Bryant 1942; Easterla 1973a, 1973b); 
Sierra Vieja Mountains, Texas (Blair and Miller 1949; specimens in collec- 
tion at Texas Tech University); Davis Mountains, Texas (Blair 1940; speci- 
mens in collection at Texas Tech University); northwestern Chihuahua, 
Mexico (Anderson 1972); Sacramento Mountains, above 5000 ft, New 
Mexico (Findley et al. 1975); Tularosa Basin below 5000 ft, New Mexico 
(Blair 1941 ; Findley et al. 1975); Lubbock County, Texas (Bailey 1905; Davis 
1974; specimens in collection at Texas Tech University). The Big Bend 
National Park, Sierra Vieja Mountains, and Davis Mountains represent 



314 GENOWAYS ET AL. 



TABLE 2. Species of mammals occurring in selected geographic areas of Texas, New Mexico, 
and Chihuahua. A plus sign indicates that the species has been recorded from the given area. 
Records are taken from literature cited in text and specimens deposited in The Museum of 
Texas Tech University. 



■a ■ 2 £ a 

J 1 £ I I " ! * d r ° 






J! O 



> a s Hi 



« 



2 



Species g « eg Q Z ^ H j_ 

Didelphis virginiana + + 

Notiosorex crawfordi + + + + 

Cryptotis parva + 

Sore* ttanus + 

5o re jc vagrans + 

Mormoops megalophylla + + + 

Leptonyctehs nivalis + 

Myotis auriculus + 

Myotis californicus + + + + + 

Myotis leibii + + + + + 

Myotis lucifugus + + 

Myotis thysanodes + + + + + + 

Myotis velifer + + + + + 

Myotis volans + + + + + + + 

Myotis yumanensis + + + + 

Pipistrellus Hesperus + + + + + + 

Lasiurus borealis + + + + 

Lasiurus cinereus + + + + + + + 

Lasionycteris noctivagans + + + + 

Eptesicus fuscus + + + + + + 

Plecotus phyllotis + 

Plecotus townsendii + + + + + 

Euderma macula turn + 

Antrozous pallidus + + + + + + 

Tadarida brasiliensis + + + + + + + + 

Tadarida femorosacca + 

Tadarida macrotis + + + + + + 

Eumops perotis + 

Sylvilagus audubonii + + + + + + + 

Sylvilagus floridanus + + + + + + + 

Lepus californicus + + + + + + + 

Lepus collotis + 

Eutamias canipes + + 

Eutamias dorsalis + 

Eutamias minimus + 

Ammospermophilus interpres + + + + 

Spermophilus mexicanus + + 

Sperrt.ophilus spilosoma + + + + + + 

Spermophilus tridecemlineatus + + 



MAMMALS 315 



TABLE 2. (continued) 



Species 





ha 
08 

fta 


t/~ 




OS 


3 










s 


C 

'3 


'3 




01 

H 


08 

3 




'3 






H 






c 
















c 

3 -* 

s ft, 

3 o 


C 

© 

"3 
Z 

■o 
e 

aa 


3 
O 

> 


08 
X 

H 


'3 
c 

3 
O 

OS 


U 

s 

1 

ha 
O 


o 
'J? 


3 

o o 

^ X 
O 8> 

c S 


OS 

ca 

OS 
O 

ha 

"5 


o 

X 

0) 

Z 


c 

3 
O 

5 

O 

JQ 
X) 

3 


U 


PQ 


tf) 




Q 


Z 




cfl 


H 




-J 



+ + + + 



Spermophilus variegatus 

Cynomys ludovicianus 

Tamiasciurus hudsonicus 

Sciurus nayaritensis 

Thomomys bottae 

Thomomys umbrinus 

Pappogeomys castanops 

Geomys arenarius 

Geomys bursarius 

Castor canadensis 

Perognathus flavescens 

Perognathus flavus 

Perognathus hispidus 

Perognathus intermedius 

Perognathus nelsoni 

Perognathus penicillatus 

Dipodomys merriami + + + + + 

Dipodomys nelsoni + 

Dipodomys ordii + + + + 

Dipodomys spectabilis •> + + + 

Reithrodontomys fulvescens + + + 

Reithrodontomys megalotis + + + + + 

Reithrodontomys montanus + + 

Peromyscus boylii + + + + + 

Peromyscus difficilis + + 

Peromyscus eremicus + + + + + 

Peromyscus leucopus + + + + 

Peromyscus maniculatus + + + + + + 

Peromyscus pectoralis + + + + 

Peromyscus truei + + 

Baiomys taylori + 

Onychomys leucogaster + + + 

Onychomys torridus + + + + 

Sigmodon fulviventer + 

Sigmodon hispidus + + + 

Sigmodon ochrognathus + + + 

Neotoma albigula + + + + + + 

Neotoma mexicana + + + + 

Neotoma micropus + + + + 

Microtus longicaudus + 

Microtus mexicanus + + 

Zapus princeps + 



316 GENOWAYS ET AL. 

TABLE 2. (continued) 



Species 


C 

"5 

is 

11 

3 o 
es s 

"2 * 
§z 


2t 
SB 
Cu 

13 

e 

« 
Z 

"3 
s 
c> 
ffi 

ex 

£ 


'3 
s 

3 
O 

m 
"37 

> s 

es x 

35 


es 
a 

H 

c 
'3 

s 

3 
O 

'> 

es 

Q 


es 
3 
-C 
es 
3 
£ 

U 

c 

aj 

tS o 

J * 

is 

z 


c 
*5j 

Is 

at 

is 
s z 


3 C 

'5 .Si 
,2 * 

es £ 

M 

1 ^ 

H 


es 
gj 

H 

s>^ 

C 
3 
O 
CJ 

u 

o 

3 


Erethizon dorsatum 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


Canis latrans 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


Canis lupus 


+ 


+ 




+ 




+ 






Urocyon cinereoargenteus 


+ 


+ 


+ 


+ 


+ 


+ 




+ 


Vulpes macrotis 




+ 




+ 


+ 




+ 




Vulpes velox 
















+ 


Vulpes vulpes 
















+ 


Ursus americanus 


+ 


+ 




+ 




+ 






Ursus arctos 


+ 






+ 


+ 








Bassariscus astutus 


+ 


+ 


+ 


+ 




+ 






Procyon lotor 


+ 


+ 


+ 


+ 


+ 


+ 




+ 


Mustela frenata 


+ 










+ 




+ 


Spilogale putorius 


+ 


+ 


+ 


+ 


+ 




+ 


+ 


Mephitis macroura 




+ 


+ 


+ 


+ 








Mephitis mephitis 


+ 


+ 


+ 


+ 


+ 




+ 


+ 


Conepatus mesoleucus 


+ 


+ 


+ 


+ 










Taxidea taxus 


+ 


+ 


+ 


+ 


+ 






+ 


Felis concolor 


+ 


+ 


+ 


+ 


+ 


+ 


+ 




Felis rufus 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


Odocoileus hemionus 


+ 


+ 


+ 


+ 


+ 


+ 


+ 




Odocoileus virginianus 


+ 


+ 




+ 


+ 


+ 




+ 


Cervus elaphus 


+ 
















Dicotyles tajacu 




+ 


+ 


+ 










Antilocapra americana 


+ 


+ 


+ 


+ 


+ 






+ 


Bison bison 


+ 






+ 


+ 






+ 


Ovis canadensis 


+ 


+ 






+ 









other mountain ranges in Trans-Pecos Texas. Northwestern Chihuahua is 
a part of the Chihuahuan Desert south of the Rio Grande. The Tularosa 
Basin is a low-lying desert region which is a northern extension of the Chi- 
huahuan Desert and contains the White Sands National Monument. The 
Sacramento Mountains are the southern extension of the Rocky Moun- 
tains in New Mexico which most closely approaches the Guadalupe Moun- 
tains. Lubbock County, Texas, is located in the southern Great Plains. 

The total mammalian fauna of the Guadalupe Mountains National Park 
shows the highest similarity with that occurring in the Davis Mountains, 
Texas (Table 3). High similarity is also shown to the total mammalian fauna 



MAMMALS 



317 



11 

^ .5 



U 45 <„ 



o c 

42 0> 



Z 



S u 
s ^ 
2 m 



3. c 

»*-l (11 



rt 



8> 



I 5 

<-■ 60 



•2 * 

<u o 

3 «, 

OS l_ 

cd 42 

£ 6 

E 2 



If 



rt 



<u 3 

4= <U 

+-> <U 

H 2 

<D *^ 

Oh <U 

2 3, 



! O 
60 
ed 

•3 

! £ £ 

B 3 £ 

£ a, S 

3 45 0> 

e +■> x 

8 e s 

<« 3^ S 

§ 2 £ 

o r 

PQ 



a 



2 £ 



o 


>, 






42' 


■M 

3 




4= 


3 


u 


3 


o 


H 


-J 





o 


ed 




.y 


s/5 

c 


_c 


9) 


J 




s 


3 


C2 


Z 



E 

a 

o 


c 
= 
o 




rt 

K 


2 


z 


C 








« 




as 


3 

■s 

3 


o 
o 

'3 


5 £ 


s 



i° 



5 h 



.2 c 

h!T *3 "5 

2 5 S 



= a, M 
a> _ ca 

ed 

z 



1 i 

3 O 

a s 



ed X 

C a> 
O H 



oc 
in 




o 


d 






-2 -2 



c 

O W3 

cd X 

Z H 

T3 „ 

3 M 



X 


o 


V3 

X 


C/3 


X 


a) 


VI 


a> 


45 


V 


H 


Q 


H 


o 
Z 


s 



o o 

§ .a 

3 S 

E * 

S3 £ 



If 

3 

H 



3 



318 GENOWAYS ET AL. 

of the Big Bend National Park and Sierra Vieja Mountains, Texas. Inter- 
mediate similarity values were obtained with northwestern Chihuahua and 
the Tularosa Basin, New Mexico, and the lowest similarity values were 
found to be with the mammalian faunas of the Sacramento Mountains, New 
Mexico, and Lubbock County, Texas. A clustering of the similarity coeffi- 
cients (Fig. 9) shows the four Trans-Pecos mountain ranges in one cluster, 
with the Guadalupe Mountains National Park being the most distinct of the 
group. The remaining four geographic areas form a series with decreasing 
similarity to these mountains — northwestern Chihuahua, Tularosa Basin, 
Lubbock County, and Sacramento Mountains. 



c 



Guadalupe Mountains National Park, Texas 
Big Bend National Park, Texas 
Davis Mountains, Texas 
Sierra Vieja Mountains, Texas 
Northwestern Chihuahua, Mexico 
Tularosa Basin, New Mexico 
Lubbock County, Texas 
Sacramento Mountains, New Mexico 



48.3 58.3 73.3 83.3 

Fig. 9. Phenogram resulting from the clustering (UPGMA) of Burt Coefficients of 
Similarity among the total mammalian faunas of selected geographic areas in 
Texas, New Mexico, and Chihuahua as given in Table 3. 



These results indicate that the total mammalian fauna of the Guadalupe 
Mountains National Park should be considered most closely related to that 
of other montane regions of Trans-Pecos Texas. The mammalian fauna is 

most distinct from those of the Sacramento Mountains and Lubbock 
County but 23 and 30 species, respectively, are shared between these areas 
and the Guadalupe Mountains National Park. 



Guadalupe Mountains National Park, Texas 
Tularosa Basin, New Mexico 
Big Bend National Park, Texas 
Davis Mountains, Texas 
Sierra Vieja Mountains, Texas 
Northwestern Chihuahua, Mexico 
Lubbock County, Texas 
Sacramento Mountains, New Mexico 



^ 



39.2 53.2 74.2 88.2 

Fig.10. Phenogram resulting from the clustering (UPGMA) of Burt Coefficients of 
Similarity among the rodent faunas of selected geographic areas in Texas, New 
Mexico, and Chihuahua as given in Table 4. 



MAMMALS 



319 



w 

pO 


c 




.O 


9 


0) 


3 





H 


-J 


u 





a 




_o 





_c 


a> 


,§ 


<n 


s 


3 


02 


Z 


o 


sS 


C 


0* 


e 




E 


E 


2 


O 


3 


l* 




2 


0) 

Z 


C 






»- 






B 


C3 






~ 





a> 


1 


_o 


£ 




'7 




H5 


0) 





U 





•- eg cs 



> 


'3 


v. 
CO 

X 
(D 

H 


>- 
>- 
0) 


c 

9 



W 


2 




-a 


3 




e 


0- 


'.o 


S3 


■a 


03 


60 


c 



3J 


c5 


£3 





a c w 

13 -2 — 4 
-S c « * 

~ 5 E 0> 

§ O.OH 

O S ts 
z 



! 



.5 « 

03 H 

I* 

5 £ 









to 


o 


o 


d 


d 



e 

O vi 
"■£5 ^ 



-^ « 



3 O 

03 Z 

3 

o 



H -5? 
m > 



W3 



03- D 



U o 

e ^ 

-E <U 

« Z 

o 

Z 



o 

c 

£ 
u 


X 
Z 


5/5 

03 
« 

03 

v> 
O 
03 


Z 


o 
U 

U 

o 

JD 

4= 


O 

H 


C3 




3 




3 




OO 




H 




J 





320 



GENOWAYS ET AL. 



TABLE 5. Species of mammals occurring in the Guadalupe Mountains National Park, Texas 
listed according to faunal units as described by Hoffmann and Jones (1970:364-365). 



Southwest 
Myotis californicus 
Myotis thysanodes 
Myotis ve lifer 
Pipistrellus Hesperus 
Antrozous pallidus 
Sylvilagus audubonii 
Lepus californicus 
Ammospermophilus interpres 
Spermophilus spilosoma 
Spermophilus variegatus 
Pappogeomys castanops 
Perognathus flavus 
Perognathus inter me dius 



Myotis volans 
Plecotus townsendii 
Eutamias canipes 
Thomomys bottae 



Cynomys ludovicianus 
Perognathus hispidus 



Tadarida brasiliensis 
Tadarida macrotis 



species (26) 

Perognathus penicillatus 
Dipodomys merriami 
Dipodomys spectabilis 
Reithrodontomys megalotis 
Peromyscus boylii 
Peromyscus difftcilis 
Peromyscus eremicus 
Peromyscus pectoralis 
Peromyscus truei 
Onychomys torridus 
Neotoma albigula 
Bassariscus astutus 
Conepatus mesoleucus 



Montane species (7) 



Neotoma mexicana 
Microtus mexicanus 
Ovis canadensis 



Steppe species (4) 



Dipodomys ordii 
Neotoma micropus 



Southern species (3) 



Sigmodon hispidus 



Sylvilagus floridanus 



Deciduous forest species (2) 

Peromyscus leucopus 



Myotis leibii 
Eptesicus fuscus 
Lasiurus cinereus 
Lasionycteris noctivagans 
Peromyscus maniculatus 
Erethizon dorsatum 
Canis latrans 
Canis lupus 

Urocyon cinereoargenteus 
Ursus americanus 
Ursus arctos 
Procyon lotor 



Widespread species (23) 

Mustela frenata 
Spilogale putorius 
Mephitis mephitis 
Taxidea taxus 
Eelis concolor 
Felis rufus 

Odocoileus hemionus 
Odocoileus virginianus 
Cervus elaphus 
Antilocapra americana 
Bison bison 



We also compared these same geographic areas using only their rodei I 
faunas. Rodents may be somewhat better indicators of faunal resemblanc 5 



MAMMALS 321 




Fig. 11. Superimposed geographic distributions of sciurid, heteromyid, and geo- 
myid rodents with affinities to the Southwest that occur in the Guadalupe Mountains 
National Park. 



because they are less vagile, are highly affected by the environment, and gen- 
erally do not have widespread geographic ranges. This changes the relation- 
ships among the areas slightly. The rodent fauna of the Guadalupe Moun- 
tains National Park shows the highest similarity to the rodent fauna of the 
Tularosa Basin (Table 4). The Tularosa Basin is a desert area lying to the 
northwest of the lowland areas of the western portion of the park. Similarity 
remains high with the Davis Mountains. 

Intermediate similarity values were obtained between the rodent faunas of 
the Big Bend National Park, Sierra Vieja Mountains, and northwestern 



322 GENOWAYS ET AL. 




Fig. 12. Superimposed geographic distributions of cricetid rodents with affinities to 
the Southwest that occur in the Guadalupe Mountains National Park. 



Chihuahua. These more southern areas lack montane and grassland forms 
such as Microtus mexicanus and Eutamias canipes that are present in the 
park and the Guadalupe Mountains National Park lacks southern arid- 
adapted species such as Perognathus nelsoni. The lowest similarity coeffi- 
cients were obtained with the Sacramento Mountains, which lacks the desert 
species, and Lubbock County, which lacks the desert and montane species. 
Clustering of these similarity values for the rodent fauna (Fig. 10) shows 
the Guadalupe Mountains closely clustered with the Tularosa Basin and in a 
major cluster with the three Trans-Pecos mountain ranges (Big Bend 
National Park, Davis Mountains, and Sierra Vieja Mountains). This cluster 



MAMMALS 323 




Fig. 13. Superimposed geographic distributions of chiropterans, lagomorphs, and 
carnivores with affinities to the Southwest that occur in the Guadalupe Mountains 
National Park. 



is progressively further from rodent faunas of the other three areas — north- 
western Chihuahua, Lubbock County, and Sacramento Mountains. 

The Guadalupe Mountains National Park's rodent fauna is dominated by 
desert-adapted species; however, it does contain some unique species. The 
total rodent fauna shows a rather distant relationship to a true grassland 
rodent fauna and even less with a true montane rodent fauna. 

The 65 species of mammals native to the Guadalupe Mountains National 
Park discussed in this account are from six faunal units (Table 5) as described 
by Hoffmann and Jones (1970:364-365). The faunal units represented are 



324 



GENOWAYS ET AL. 



Southwest (26 species), Montane (7), Steppe (4), South (3), Deciduous 
Forest (2), and Widespread Species (23). Widespread Species are those with 
broad geographic ranges and are of little value in determining the relation- 
ships of a fauna. 

Figures 11-13 show the superimposed geographic ranges of the species 
believed to have Southwestern affinities. The sciurid, heteromyid, and 
geomyid rodents occurring in the park have a center of distribution in the 
Chihuahuan and eastern portion of the Sonoran deserts. The cricetid 
rodents of the park are centered basically on the Chihuahuan Desert and the 




Fig. 14. Superimposed geographic distributions of mammals with montane af- 
finities that occur in the Guadalupe Mountains National Park. 



MAMMALS 



325 



nonrodent species center on the Chihuahuan and eastern Sonoran deserts. 
The seven montane species shown in Fig. 14 have a center of distribution that 
includes the southern Rocky Mountains and southwestern desert ranges. 
Although these are animals from high elevations, they represent primarily 
the southwestern extension of this faunal unit. The four Steppe species from 
the park (Fig. 15) have a distributional center on the southern Great Plains 
and northern Chihuahuan Desert. In the desert regions, these species are 
occurring basically in desert grasslands as we have seen in the park. 
The species of mammals from the South are basically tropical and sub- 




Fig. 15. Superimposed geographic distributions of mammals with Steppe affinities 
that occur in the Guadalupe Mountains National Park. 



326 GENOWAYS ET AL. 

tropical in distribution. The two species of Tadarida are migratory and occur 
in the park only during the summer months; they overwinter in central 
Mexico or farther south. The hispid cotton rat is a species that is still 
expanding its geographic range northward. The two species with affinities to 
the deciduous forest (Sylvilagus floridanus and Peromyscus leucopus) have 
relatively wide ranges but almost always occur in forested areas. 

The mammalian fauna of Guadalupe Mountains National Park is pre- 
dominately Southwestern in affinities, with the Chihuahuan Desert forms 
being the chief component. The montane faunal unit, although third in num- 
ber of species (seven), contains relatively few of the total mammalian species 
in the park. However, it was for the preservation of this unique faunal com- 
ponent, which occupies a mountaintop island, for which the Guadalupe 
Mountains National Park was established. 

Recommendations 

This survey must be considered as a starting point, which supplies base- 
line data, and is not an end in itself. With this in mind, we have submitted the 
following recommendations to the National Park Service for future work 
and development of the park. 

1 . An inventory of the larger mammals in the park should be undertaken. 
These animals (such as the elk, mule deer, bear, coyote, bobcat, mountain 
lion, and fox) have an important role in the ecosystem. It is essential to the 
success of resource management in the park that more information is 
gathered concerning these animals. 

2. It is essential that the status of the elk herd be fully investigated. Knowl- 
edge of their food habits, age structure of the herd, reproductive success, 
herd movements, and water stress is needed. 

3. A program of continuous monitoring of mammalian populations with 
scheduled periodic censuses should be established. A combination of grid 
trapping supplemented With general trapping with live traps is recom- 
mended for detecting population trends in small mammals. For monitoring 
the larger mammals a combination of aerial census, field observations, and 
fecal pellet group analysis could be used. The following sites are recom- 
mended for periodic censusing: Upper Dog Canyon Ranger Station; The 
Bowl; Nipple Hill; Pratt Lodge — McKittrick Canyon; entrance to Williams 
Ranch Road; Lewis Well; Crossroads at the north end of the Patterson Hills. 

4. Special efforts must be made during census procedures to monitor the 
status of the species that are rare in the park: Microtus mexicanus 
guadalupensis; Sylvilagus floridanus robustus; Thomomys bottae 
guadalupensis; Eutamias canipes canipes; Ammospermophilus interpres. 
Careful monitoring of these mammals may prevent their loss from the park. 

5. The grid now established in Upper Dog Canyon should be made 
permanent. Additional permanent population grids should be established at 
the following sites: The Bowl; between Nipple Hill and Choza Spring or Pine 
Springs Canyon; Williams Ranch House; Lewis Well. Periodic live trapping 



MAMMALS 327 

on these grids would yield much valuable data and detect population trends. 

6. The following areas are considered to be the most biologically 
significant from the standpoint of mammals in Guadalupe Mountains 
National Park: Upper Dog Canyon; The Bowl; Nipple Hill area; Lewis Well; 
McKittrick Canyon; every spring and waterhole in the park. 

7. The zones of contact between Neotoma albigula and N. mexicana in 
Upper Dog Canyon and between N. albigula and N. micropus north of the 
Patterson Hills should be monitored periodically to note any shifts in the 
zones. Such shifts may indicate changing environmental conditions. 

8. Data from periodic censuses should be used to test the computer model 
under development for remote sensing of the park. Remote sensing should 
prove especially useful in the monitoring of vole habitat. 

9. A special effort should be made to determine if the following mammals 
occur within the park boundaries: Notiosorex crawfordi\ Myotis evotis; 
Euderma maculatum; Lasiurus borealis; Spermophilus mexicanus; 
Tamiasciurus hudsonicus; Onychomys leucogaster\ Vulpes macrotis; 
Dicotyles tajacu. 

10. High priority should be given to repairing or rebuilding the boundary 
fences on the south and west sides of the park. During the mammal survey, a 
large number of cattle have been observed within the park boundaries. The 
number of trespassing cattle has increased in the last two years and they have 
recently been observed at the west base of the Guadalupe escarpment. This 
area of the park recovers from grazing very slowly in the total absence of live- 
stock; therefore, it is imperative that the fences be repaired and the cattle 
removed. 

1 1 . One of the real joys for tourists, especially children, is seeing deer and 
other wildlife at Pine Springs Campground and Upper Dog Canyon camp- 
site. Several times we have observed unleashed dogs chasing deer and other 
mammals away. Every effort should be made to enforce the existing leash 
law at the park. A little thoughtfulness on the part of pet owners will allow 
everyone to share the experience of viewing some of the park's wildlife. This 
is not a criticism of the current park personnel as we have observed them 
enforcing the leash laws. Perhaps, pets should be excluded from the park. If 
people want to take their pets on vacation, they can stay at private camp- 
grounds and leave their pets there while visiting the National Park. 

12. Most of the endangered species that occur in Guadalupe Mountains 
National Park probably occur in Carlsbad Caverns National Park as well. 
We would strongly recommend that a mammal survey be conducted in 
Carlsbad Caverns National Park. This research would provide valuable 
information which should be included in the final environmental statement 
for the master plan proposed for Carlsbad Caverns National Park. If the 
endangered species are present in Carlsbad Caverns National Park, a 
management program consistent in both parks could be developed which 
would increase the possibilities of preserving the species for future genera- 
tions. 



328 GENOWAYS ET AL. 

LITERATURE CITED 

Allen, L. A. 1877. History of the American bison. Bison americanus. Annu. Rep. 

U.S. Geol. Surv.for 1875 9:443-590. 
Anderson, A. E., W. A. Snyder, and G. W. Brown. 1965. Stomach content 

analyses related to condition in mule deer, Guadalupe Mountains, New Mexico. J. 

Wildl. Manage. 29:352 366. 
1970. Indices of reproduction and survival in female mule deer, Guadalupe 

Mountains, New Mexico. Southwest. Nat. 15:29-36. 
Anderson, S. 1972. Mammals of Chihuahua: taxonomy and distribution. Bull. Am. 

Mus. Nat. Hist. 148:149-410. 
August, P. V., J. W. Clarke, M. H. McGAUGRand R. L. Packard 1977. 

Demographic patterns of small mammals: a possible use in impact assessment. 

This volume. 
Bailey, V. 1902a. Synopsis of the North American species of Sigmodon. Proc. Biol. 

Soc. Wash. 15:101-116. 
1902b. Seven new mammals from western Texas. Proc. Biol. Soc. Wash. 

15:117-120. 

1905. Biological survey of Texas. North Am. Fauna 25:1-222. 

1932. Mammals of New Mexico. North Am. Fauna 53.1-412. 



Bailey, V., and C. C. S perry. 1929. Life history and habits of the grasshopper 

mice, genus Onychomys. U.S. Dep. Agric, Tech. Bull. 145:1-19. 
Baker, R. J. 1970. Karyotypic trends in bats. Pages 65-96 in W. A. Wimsatt, ed. 

Biology of Bats, Academic Press, New York, 406 pp. 
Barbour, R. W., and W. H. Davis. 1969. Bats of America. Univ. Press Kentucky, 

Lexington, 286 pp. 
Birney, E. C. 1973. Systematics of three species of woodrats (genus Neotoma) in 

central North America. Misc. Publ. Mus. Nat. Hist., Univ. Kans. 58:1-173. 
Blair, W. F. 1940. A contribution to the ecology and faunal relationships of the 

mammals of the Davis Mountains region, southwestern Texas. Misc. Publ. Mus. 

Zool., Univ. Mich. 46:1-39. 
1941. Annotated list of mammals of theTularosa Basin, New Mexico. Am. 

Midi. Nat. 26:218-229. 
Blair, W. F., and C. E. M iller, JR. 1949. The mammals of the Sierra Vieja region, 

southwestern Texas, with remarks on the biogeographic position of the region. 

Tex. J. Sci. 1:67-92. 
Bogan, M. A. 1975. Geographic variation in Myotis californicus in the south- 
western United States and Mexico. Wildl. Res. Rep., Fish and Wildlife Serv. 

3:1-31. 
Borell, A. E., and M. D. Bryant. 1942. Mammals of the Big Bend area of Texas. 

Univ. Calif. Publ. Zool. 48:1-62. 
Buechner, H. K. 1950. Life history, ecology, and range use of the pronghorn 

antelope in Trans-Pecos Texas. Am. Midi. Nat. 43:257-354. 
COCKRUM, E. L. 1969. Migration in the guano bat, Tadarida brasiliensis. Pages 

303-336 iTi J. K. Jones, Jr., ed. Contributions in mammalogy. Misc. Publ. Mus. 

Nat. Hist., Univ. Kans. 51:1-428. 
COLE, H. F.,and H. H.Wolfe. 1970. Laboratory evaluation of aggressive behavior 

of the grasshopper mouse {Onychomys). J. Pharm. Sci. 59:969-971. 
Cornely. J. E. 1976. Checklist of Mammals, Guadalupe Mountains National Park. 

Texas. Texas Tech Press, Lubbock, Texas, 4 pp. 



MAMMALS 329 

1977. Ecological distribution of woodrats (genus Neotoma) in Guadalupe 

Mountains National Park, Texas. This volume. 
Cowan, I. McT. 1956. What and where are the mule and black-tailed deer? Pages 

334-359 in W. P. Taylor, ed. The Deer of North America. Stackpole Co., 

Harrisburg, Pa., 668 pp. 
CyrM. 1972. Predatory behavior of the grasshopper mouse, Onychomys torridus. 

Ph. D. Thesis, Univ. California Los Angeles. 
Davis, W. B. 1940. Mammals of the Guadalupe Mountains of western Texas. 

Occas. Pap. Mus. Zool, Louisiana State Univ. 7:69-84. 
1960. The mammals of Texas. Bull. Game and Fish Commission, Austin, 

Texas 41:1-252. 
1974. The mammals of Texas. Bull. Tex. Parks and Wildl. Dep. 



41(revised): 1-294. 
Davis, W. B., and J. L. Robertson, Jr. 1944. The mammals of Culberson Coun- 
ty, Texas. /. Mammal. 25:254-273. 
Davis, W. B., and W. P. Taylor. 1939. The bighorn sheep of Texas. J. Mammal. 

20:440-455. 
DlERSlNG, V. E., and D. F. Hoffmeister. 1974. The rock mouse, Peromyscus dif- 

ficilis, in western Texas. Southwest. Nat. 19:213. 
Easterla, D. A. 1973a. Ecology of the 18 species of Chiroptera at Big Bend 

National Park, Texas. Part I. Northwest Missouri State Univ. Studies 34:1-53. 
__. 1973b. Ecology of the 18 species of Chiroptera at Big Bend National Park, 

Texas. Part II. Northwest Missouri State Univ. Studies 34:54-165. 
Engels, W. L. 1936. Distribution of races of the brown bat (Eptesicus) in western 

North America. Am. Midi. Nat. 17:653-660. 
Findley, J. S., and G. L. Traut. 1970. Geographic variation in Pipistrellus 

hesperus. J. Mammal. 51:741-765. 
Findley, J. S., A. H. Harris, D. E. Wilson, and C. Jones 1975. Mammals of 

New Mexico. Univ. New Mexico Press, Albuquerque, 360 pp. 
Fleharty, E. D. 1960. The status of the gray-necked chipmunk in New Mexico. J. 

Mammal. 41:235-242. 

Gehlbach, F. R. 1967. Vegetation of the Guadalupe escarpment, New Mex- 
ico-Texas. Ecology 48:404-419. 

Undated. Description and development of the living landscape of Guada- 
lupe Mountains National Park Texas. Rep. National Park Service, U.S. Dep. In- 
ter., Santa Fe, New Mexico. 

Glass, B. P. 1947. Geographic variation in Perognathus hispidus. J. Mammal. 
28:174-179. 

Glass, B. P., and R. J. Baker. 1968. The status of the name Myotis subulatus 
Say. Proc. Biol. Soc. Wash. 81:257-260. 

Goldman, E. A. 1936. New pocket gophers of the genus Thomomys. J. Wash. 
Acad. Sci. 26:111-120. 

1944. Classification of wolves. Pages 389-507 in S. P. Young and E. A. 

Goldman eds. The Wolves of North America. American Wildlife Institute, 
Washington, D.C. 636 pp. 

1946. Classification of the races of the puma. Pages 175-302 in S. P. Young 



and E. A. Goldman eds. The Puma, Mysterious American Cat. Wildlife Institute, 
Washington, D.C, 358 pp. 



330 GENOWAYS ET AL 



1950. Raccoons of North and Middle America. North Am. Fauna 



60:1-153. 

Gross, J. E. 1960. History, present, and future status of the desert bighorn sheep 
(Ovis canadensis mexicana) in the Guadalupe Mountains of southeastern New 
Mexico and northwestern Texas. Trans. Desert Bighorn Council 4:66-7 \ . 

Hall, E. R. 1951a. American weasels. Univ. Kans. Pubi, Mus. Nat. Hist. 4:1-466. 

1951b. A synopsis of the North American Lagomorpha. Univ. Kans. Pubi, 

Mus. Nat. Hist. 5:119-202. 

Hall, E. R., and K. R. Kelson. 1951. Comments on the taxonomy and geo- 
graphic distribution of some North American rabbits. Univ. Kans. Pubi, Mus. 
Nat. Hist. 5:49-58. 

Handley, C. O., Jr. 1959. A revision of American bats of the genera Euderma and 
Plecotus. Proc. U.S. Natl. Mus. 110-95-246. 

Hayward, B. J. 1970. The natural history of the cave bat, Myotis velifer. Western 
New Mexico Univ. Res. Sci. 1:1-74. 

Hoffmann, R. S., and J. K.Jones, Jr. 1970. Influence of late-glacial and post- 
glacial events on the distribution of Recent mammals on the Northern Great 
Plains. Pages 355-394 in W. Dort, Jr., and J. K. Jones, Jr., eds. Pleistocene and 
Recent environments of the central Great Plains. Univ. Press Kansas, 433 pp. 

Hoffmeister, D. F. 1951. A taxonomic and evolutionary study of the pinon 
mouse, Peromyscus truei. Illinois Biol. Monogr. 21:1-104. 

Hoffmeister, D. F., and L. de la Torre. 1961. Geographic variation in the 
mouse Peromyscus difficilis. J. Mammal. 42:1-13. 

Hoffmeister, D. F., and M. R. Lee. 1967. Revision of the pocket mice, 
Perognathus penicillatus. J. Mammal. 48:361-380. 

Hollister, N. 1916. A systematic account of the prairie dogs. North Am. Fauna 
40:1-37. 

Horner, B. E., J. M. Taylor, and H. A. Padykula. 1964. Food habits and gas- 
tric morphology of the grasshopper mouse. J. Mammal. 49:644-660. 

Howell, A. H. 1914. Revision of the American harvest mice (genus Reithrodon- 
tomys). North Am. Fauna 36:1-97. 

1938. Revision of the North American ground squirrels. North Am. Fauna 

56:1-256. 

HUSSON, A. M. 1962. The bats of Suriname. Zool. Verhand. 58:1-282. 

Jackson, H. H. T. 1951. Classification of the races of the coyote. Pages 229-315 in 
S. P. Young and H. H. T. Jackson, eds. The Clever Coyote. Stackpole Co., 
Harrisburg, Pa., 411 pp. 

Kellogg, R. 1956. What and where are the whitetails? Pages 31-55 in W. P. 
Taylor, ed. The Deer of North America. Stackpole Co., Harrisburg, Pa., 668 pp. 

Kittams, W. H., S. L. Evans, and D. C. Cooke 1977. Food habits of mule deer 
on foothills of Carlsbad Caverns National Park. This volume. 

La Val, R. K. 1973. Occurrence, ecological distribution, and relative abundance of 
bats in McKittrick Canyon, Culberson Co., Texas. Southwest. Nat. 17:357-364. 

Lidicker, W. Z., Jr. 1960. An analysis of intraspecific variation in the kangaroo 
rat, Dipodomys merriami. Univ. Calif. Pubi. Zool. 67:125-218. 

Long, C. A. 1972. Taxonomic revision of the North American badger, Taxidea 
taxus. J. Mammal. 53:725-759. 

McCarty, R. 1975. Onychomys torridus. Mamm. Species 59:1-5. 



MAMMALS 331 

McCullough, D. R. 1969. The Tule elk: its history, behavior, and ecology. 

Univer. Calif. Publ. Zool. 88:1-191. 
Mead, R. A. 1967. Age determination in the spotted skunk. J. Mammal. 

48:606-616. 
1968a. Reproduction in eastern forms of the spotted skunk (genus 

Spilogale). J. Zool. 156:119-136. 
1968b. Reproduction in western forms of the spotted skunk (genus 



Spilogale). J. Mammal. 49:373-390. 
M earns, E. A. 1896. Preliminary diagnoses of new mammals from the Mexican 

border of the United States. Proc. U.S. Natl. Mus. 18:443-447. 
Mollhagen, T., and R. H. Baker, 1972. Myotis volans interior Miller in Knox 

County, Texas, Southwest. Nat. 17:97. 
Murie, O. J. 1951. The Elk of North America. Stackpole, Co., Harrisburg, Pa., 

376 pp. 

Nelson, E. W. 1909. The rabbits of North America. North Am. Fauna 29:1-314. 
1925. Status of the pronghorned antelope, 1922-1924. Bull. U.S. Dep. 

Agric. 1346: -64. 
O'Connell, M. A. 1977. Co-existence between two species of kangaroo rate (genus 

Dipodomys) in the Guadalupe Mountains National Park, Texas. This volume. 
Pizzimenti, J. J. 1975. Evolution of the prairie dog genus Cynomys. Occas. Pap. 

Mus. Nat. Hist., Univ. Kansas 29:1-73. 
Russell, R. J. 1968. Revision of pocket gophers of the genus Pappogeomys. Univ. 

Kans. Publ, Mus. Nat. Hist. 16:581-776. 
Schmidly, D. J. 1972. Geographic variation in the white-ankled mouse, 

Peromyscus pectoralis. Southwest. Nat. 17:113-138. 
1973. Geographic variation and taxonomy of Peromyscus boylii from 

Mexico and the southern United States. J. Mammal. 54:111-130. 
Setzer, H. W. 1949. Subspeciation in the kangaroo rat, Dipodomys ordii. Univ. 

Kans. Publ, Mus. Nat. Hist. 1:473-573. 
Snyder, W. A. 1961. A chemical analysis of 13 major deer forage plants from the 

Guadalupe Mountains of New Mexico and their adequacy for maintaining deer. 

M.S. Thesis, New Mexico State Univ., Las Cruces, 58 pp. 
Van Gelder, R. G. 1959. A taxonomic revision of the spotted skunks (genus 

Spilogale). Bull. Am. Mus. Nat. Hist. 117:229-392. 
Warnock, B. H. Undated. Plant communities of the Guadalupe Mountains in 

Texas and nearby Carlsbad Caverns National Park. Rep. submitted to Carlsbad 

Caverns National Park, Carlsbad, New Mexico. 
Wilhelm, D. E. 1977. Status of the Guadalupe Mountains vole, Microtus mex- 

icanus guadalupensis. This volume. 
Wilson, D. E. 1973. The systematic status of Perognathus merriami Allen. Proc. 

Biol Soc. Wash. 86:175-192. 
Wright, G. M., and B. H. Thompson. 1934. Wildlife management in the national 

parks. Fauna Natl. Parks 2:viii, 1-142. 



332 GENOWAYS ET AL. 

ACKNOWLEDGMENTS 

We most gratefully acknowledge the field assistance of Dallas E. Wilhelm 
and Margaret A. O'Connell. Without their assistance, this project would not 
have been completed. Our work in the park was facilitated greatly by the 
following park personnel: Donald A. Dayton; John Chapman; Gary M. 
Ahlstrand; Philip Van Cleave; Roger Reisch. We acknowledge the Depart- 
ment of Biological Sciences and The Museum of Texas Tech University for 
supporting most of the laboratory phases of this study. Field portions of this 
work were supported by a contract with the National Park Service 
(CX700040145), administered by Mr. Roland H. Wauer, Chief Scientist, 
Southwest Region, Santa Fe, New Mexico. Mr. Stephen L. Williams pre- 
pared the line drawings. We would like to thank Dr. David J. Schmidly, 
Texas Cooperative Wildlife Collection, Texas A&M University, and Dr. 
Don E. Wilson, National Bird and Mammal Laboratory, National Museum 
of Natural History, for allowing us to examine specimens in their care. 



Demographic Patterns of Small 
Mammals: A Possible Use in 
Impact Assessment 



PETER V. AUGUST, JOHN W. CLARKE, M. HOUSTON 
MCGAUGH and ROBERT L. PACKARD, Texas Tech 
University and The Museum, Lubbock 

Valuable inventory studies of the flora, fauna, and other related factors 
have been initiated in the Guadalupe Mountains National Park in the past 
few years. The aim of one of these has been to establish intensive study sites 
to monitor demographic patterns in small mammals. The assessment of 
these patterns may prove helpful in planning future public use of various 
areas within the park. When specific areas have been opened to visitor 
traffic, a system of monitoring the effects of human impact in these areas 
should be continued. 

Changes in vegetational composition as a result of impact have usually 
been assayed before and after human use. Also common are comparative 
studies of floral elements of an area open to human traffic with those of a 
nearby or similar area with restricted visitation (Whitson 1974). Most assess- 
ment efforts have been restricted to vegetational analyses because floristic 
structure is one of the most important and readily studied elements in a local 
ecosystem; however, there is no reason to restrict inventories to plants for 
impact use. Demographic patterns of small mammals may be important in 
the determination of the effects of human use upon a given area. 

Because of the close ecological relationship between small mammals and 
the floral components of an ecosystem, we suspect that human use will affect 
the structure and function of small mammal populations. To test this 
hypothesis, dynamics of both the plant and animal elements of an area must 
be measured before and after the disturbance. We are presently engaged in 
gathering pre-use, baseline data. In the following, we will discuss the tech- 
niques employed in measuring different parameters of small mammal popu- 
lations, demographic data collected to date, and sampling strategies for 
future censuses. 

333 



334 AUGUST ET AL. 

MATERIALS AND METHODS 

Three live-trapping grids were established in different localities in 1974 
and 1975. Each grid was designed in a fashion similar to that prescribed by 
the International Biological Program (French 1971); stations were located at 
7.5-m intervals in 12 rows by 12 columns, resulting in a grid of 144 stations, 
covering 0.81 ha. Each station was marked by a 1-m wooden stake painted 
with the row and column number for location identification. The three study 
areas are located at the following places within the park. 
Upper Dog Canyon. — The Upper Dog Canyon study site was established in 
May 1974. The grid is located in a grassy meadow, approximately one- 
fourth mile north of the ranger's cabin (elevation 1890 m). Dominant vege- 
tation of the grid includes Stipa tenuissima, Muhlenbergia repens, Juniperus 
deppeana, and Opuntia sp. (Fig. 1). Access to Dog Canyon is presently 
limited to researchers, park personnel, and backpackers. Trapping data were 
collected in 1974 and 1975. 






FIG. 1. Upper Dog Canyon study area. 

Pine Springs. — The Pine Springs study site, established in May 1975, is 
located just north of the Houser House (elevation 1 67 1 m). This location was 
chosen because it is typical of the desert-grassland vegetation of the lower 
elevations (Fig. 2). Most common grasses found on the study site are 
Bouteloua curtipendula, B. gracilis, Lycurus phleoides, Muhlenbergia 
setifolia, Artistida glauca, and Stipa neomexicana. Common perennials 



MAMMALIAN DEMOGRAPHY 335 



^ 






ftt" 


If; 1 








FIG. 2. Pine Springs study area. 

include Parthenium incanum, Opuntia imbricata, and Dasyliron sp. 
(Northington and Burgess 1977). Presently, this area receives minimal 
human traffic, but increases are expected because of close proximity to the 
Bear Canyon Trail. 

Williams Ranch. — The Williams Ranch site, established in May 1975, is on 
the west side of the Guadalupes at an elevation of 1318 m. The grid is 
approximately one-fourth mile inside the entrance gate and is bisected by a 
dirt road 7 m wide. This area is typical desert-scrub habitat dominated by 
Larrea tridentata, Prosopis glandulosa, Acacia neovericosa, Flourensia 
cernua, and Atriplex canescens (Fig. 3). In association with these 
microphyllous shrubs are the grasses, Bouteioua eripoda, Sporobolus 
airoides, and Erioneuron pulchellum (Northington and Burgess 1977). This 
area is accessible to Park visitors, on a limited basis, by a narrow dirt road. 
As the Park grows in popularity, increased visitor traffic is anticipated. This 
site should provide valuable data on the effects of increased road usage on 
small mammal populations. 

All mammal trapping was conducted in late spring and summer, and each 
trapping period was for 5 consecutive days. Sherman live traps, baited with 
rolled oats, were used throughout the study. Traps were set (one per grid sta- 
tion) in the late afternoon and examined early the following morning. In a 
few instances, traps were examined in the middle of the night to minimize the 
risk of killing animals in unusually cold weather. The following were 
recorded for each capture: station number; species and number of each (all 



336 AUGUST ET AL. 



r 






-,- '<i&i,,; 







FIG. 3. Williams Ranch study area. 

mammals were given a toe clip formula); sex; reproductive condition (testes 
scrotal or abdominal in males, vulva inactive, turgid, cornified, or pregnant 
in females); age; physical condition. These data were recorded on a 
standardized sheet developed by the International Biological Program (IBP) 
(form NREL-10). After each census period, the data were transferred to 
computer cards for storage and future analysis. 

Home ranges were calculated using the exclusive boundary strip method 
(Stickel 1954). This procedure was chosen because the home-range estimates 
produced were conservative. Also, the technique did not require a large 
number of recaptures per individual. Animals with three or more captures 
were used in the analysis. Density estimates were determined using the mini- 
mum number known alive technique (Zippin 1958). A detailed description of 
techniques employed in this study will appear in a later paper. 

RESULTS AND DISCUSSION 

Results reported herein are preliminary. One year's sampling data have 
been gathered for the Pine Springs and Williams Ranch grids, whereas data 
for 2 years have been collected for the Upper Dog Canyon grid. Year-to- 
year comparisons cannot be attempted with such minimal information. 

Upper Dog Canyon 

Six species of mammals were caught in the Upper Dog Canyon study area 
in 1974. Density estimates for 1974 in number of animals per hectare are as 



MAMMALIAN DEMOGRAPHY 337 

TABLE 1. Individuals, number of captures (in parentheses), and home ranges of rodents 
caught in the study areas. Home ranges are given only for those species with adequate num- 
ber of recaptures. 

Locality and Number of Home 

species individuals ranges un 2 ) 

Pine Springs 

Perognathus flavus 
Peromyscus maniculatus 
Peromyscus pectoralis 
Reithrodontomys megalotis 
Sigmodon hispidus 
Neotoma micropus 

Williams Ranch 

Perognathus penicillatus 
Perognathus intermedius 
Dipodomys merriami 
Dipodpmys spectabilis 
Dipodomys ordii 
Peromyscus maniculatus 
Peromyscus eremicus 
Onychomys torridus 
Neotoma micropus 
Spermophilus spilosoma 
Spermophilus mexicanus 

Upper Dog Canyon, 1974 
Peromyscus sp. 
Reithrodontomys megalotis 
Sigmodon hispidus 
Neotoma albigula 
Microtus mexicanus 

Upper Dog Canyon, 1975 
Peromyscus boylii 
Peromyscus pectoralis 
Reithrodontomys megalotis 
Sigmodon hispidus 
Neotoma albigula 
Microtus mexicanus 



follows: all rodents, 103.0; Microtus mexicanus, 1.2; Neotoma albigula, 
23.5; Sigmodon hispidus, 42.0; Reithrodontomys megalotis, 37.0; 
Peromyscus sp., 23.5. Peromyscus species identification was not made in the 
1974 sampling. Recent inspection of museum specimens suggests P. boylii 
and P. pectoralis were the species caught on the grid. See Genoways et al. 
(1977) for a discussion of the Peromyscus of Upper Dog Canyon. Home- 
range data are given in Table 1. 

Reproductive data suggest breeding occurred throughout the summer. 



3(4) 




6(17) 


416 


12(30) 


499 


KD 




6(16) 


262 


KD 




11(20) 


469 


4(12) 


732 


12(44) 


573 


3(22) 


753 


2(2) 




10(20) 


547 


7(10) 




7(27) 


1203 


15(31) 


258 


5(6) 




1(1) 




19(51) 


286 


30(46) 


511 


34(81) 


893 


17(62) 


217 


KD 




6(33) 


987 


5(22) 


382 


29(46) 


371 


11(53) 


1347 


5(15) 


262 


8(10) 


169 



338 AUGUST ET AL. 

Seventy-one percent of the rodents were sexually active in the June census, 
77%, in July, and 71%, in August. 

In 1975, six species of small mammals were caught regularly (Table 1). 
Sigmodon hispidus and Reithrodontomys megalotis accounted for 56% of 
the total captures and were likely the dominant rodents of the study grid. In 
number of animals per hectare, the following density estimates for 1975 were 
calculated: all rodents, 81.5; Microtus mexicanus, 9.9; Peromyscus boylii, 
8.6; P. pectoralis, 6.2; Reithrodontomys megalotis, 35.8; Sigmodon 
hispidus, 14.8; Neotoma albigula, 6.2. 

The average home ranges of all taxa caught are shown in Table 1. 
Microtus and Neotoma had the smallest home ranges. Because of the few 
captures, the Microtus estimate may not reflect a true home-range size. The 
small home range of Neotoma suggests they confine their movements to the 
area of their nest site. 

The reproductive data indicate rodents in the Upper Dog Canyon grid 
reproduced later in the year when compared with the other two study areas 
(Fig. 4). This probably results because principal growth of vegetation 
(primarily grasses) occurs toward the end of summer. In the lower elevation 
study areas, primary production (herbs and shrubs) begins in early to middle 
summer. Another possible explanation of this latent breeding season 
observed in Upper Dog Canyon may relate to rainfall patterns observed in 
the Guadalupes in 1975. Williams Ranch and Pine Springs received summer 



100 



D) 




C 

• «■■■* 


80 


~0 




<D 






60 


-D 




+_ 




C 


40 


<D 




<J 




i_ 




<D 


20 


CL 







June 



July 



august 



FIG. 4. Reproductive data for small mammals caught in 1975. Diagonal bars 
represent Upper Dog Canyon grid, dots represent Williams Ranch grid, and vertical 
bars represent Pine Springs grid. 



MAMMALIAN DEMOGRAPHY 339 

rains earlier than did Upper Dog Canyon. The desert-adapted annual vege- 
tation of the lower grids responds in growth more quickly to precipitation 
than grasses (Tony L. Burgess, pers. comm.). 

There is a high rate of population turnover in the Upper Dog Canyon grid, 
based on the number of animals caught both sampling years. One 
Sigmodon, two Neotoma, and one Peromyscus bcylii were caught both 
years. The Microtus population increased significantly during the second 
sampling period. In 1974, only one individual was captured, whereas eight 
were captured in 1975. Microtus mexicanus has a distribution restricted to 
grassy meadows of the higher elevations. Efforts should be made to protect 
the few areas of favorable habitat of this endemic vole. 

Pine Springs 

Of the three study areas, Pine Springs had the iowest population density. 
In number of animals per hectare, estimated densities in 1975 were as 
follows: all rodents, 35.0; Peromyscus maniculatus, 7.5; P. pectoralis, 15.0; 
Neotoma micropus, 1.2; Perognathus flavus, 3.8; Reithrodontomys mega- 
lotis, 1.2; Sigmodon hispidus, 7.5. Home ranges of Peromyscus and 
Sigmodon are shown in Table 1. Home ranges were not calculated for 
Perognathus, Reithrodontomys, and Neotoma because of a lack of suffi- 
cient recapture data. Reproductive data are given in Fig. 4. 

Williams Ranch 

This area had the highest density and greatest rodent diversity of the three 
study areas. In number of animals per hectare, estimated population 
densities in 1975 were as follows: all rodents, 95.1; Dipodomys merriami, 
16.0; D. ordii, 2.5; D. spectabilis, 3.7; Neotoma micropus, 18.5; Onychomys 
torridus, 8.6; Peromyscus eremicus, 8.6; P. maniculatus, 8.1; Perognathus 
intermedius, 4.9; P. penicillatus, 11.1; Spermophilus spilosoma, 6.2; and S. 
mexicanus, 1.2. 

Sufficient recapture data permit the estimate of home ranges for 7 of the 
11 species (Table 1). Onychomys had the largest home range of all the 
rodents. This may result from the wide-ranging foraging habits of this car- 
nivorous rodent. The remaining species are herbivores or granivores and had 
similar home-range sizes — about half that of Onychomys. Neotoma 
micropus had an extremely small home range as a result of restrictive move- 
ments about the vicinity of the nest site. Reproductive data are given in 
Fig. 4. 

FUTURE SAMPLING IN THE GUADALUPE MOUNTAINS 

Small mammal populations undergo cyclic shifts in numbers and repro- 
ductive rates (Krebs et al. 1973). These cycles are governed by extrinsic 
forces such as food availability, cover, and home site availability, as well as 
intrinsic forces such as behavioral patterns, density, and species diversity. A 
population must be sampled over a long period of time to ascertain the 



340 AUGUST ET AL. 

magnitude of cyclic oscillations. It is for this reason that data are being col- 
lected now in relatively undisturbed study areas. When these areas are 
opened for heavy visitor use, pre-use baseline information will be available 
to compare with post-impact demographic data. It is imperative that the pre- 
use data include the range of amplitudes in population densities throughout 
the seasons. This is necessary in order to minimize the possibility of 
confusing impact of human use with normal oscillation phenomena. Vegeta- 
tional components of ecosystems are generally affected first by perturba- 
tions. This, in turn, influences rodent population parameters. Care must be 
exercised in assaying visitor impact on the small mammals without first 
referring to data on the vegetational components. For this reason, col- 
lection of quantitative vegetational structure data coupled with demo- 
graphic data is needed in the future. 

REFERENCES 

French, N. R. 1971. Basic field data collection procedures for the Grassland 
Biome 1971 season. U.S. IBP Grassland Biome Tech. Rep., Colorado State 
Univ., Fort Collins, 85:1-53. 

Genoways, H. H., R. J. Baker, and J. E. Cornely. 1977. Mammals of the 
Guadalupe Mountains National Park, Texas. This volume. 

Krebs, C. J., M. S. Gaines, B. L. Keller, J. H. Myers, and R. H. Tamarin. 
1973. Population cycles in small rodents. Science 179:35-41. 

Northington, D. K., and T. L. Burgess. 1977. Summary of the vegetative zones 
of the Guadalupe Mountains National Park, Texas. This volume. 

Stickel, L. F. 1954. A comparison of certain methods of measuring ranges of 
small mammals. J. Mammal. 35:1-15. 

Whitson, P. D. 1974. The Impact of Human Use Upon the Chisos Basin and Ad- 
jacent Lands. Natl. Park Serv. Monogr. Ser. No. 4, 92 p. 

Zippin, C. 1958. The removal method of population estimations. /. Wildl. Manage. 
22:82-90. 



ACKNOWLEDGMENTS 

James Cottrell collected the field data in 1974. John Cornely, Margaret 
O'Connell, Norman Stephens, and Lloyd Logan assisted in the field work. 
We are grateful to James G. Hallett, Stuart Pimm, Kenneth G. Matocha, 
and James B. Montgomery for their suggestions during the course of the 
study. 



Population Size of Tadarida 
brasiliensis at Carlsbad Caverns 
in 1973 



J. SCOTT ALTENBACH and KENNETH N. GELUSO, 
University of New Mexico, Albuquerque 

DON E. WILSON, National Museum of Natural History, 
Washington, D.C. 

The evening outflight of the summer colony of Tadarida brasiliensis 
(Mexican free-tailed bat) from Carlsbad Cavern, Eddy County, New 
Mexico, has been an important drawing card for tourism to Carlsbad 
Caverns National Park. In recent years National Park Service personnel 
have reported a reduction in the density of the outflight, the area of the roost 
occupied by bats, and the noise level in the roost area (National Park Service 
Personnel, pers. comm.). A population size estimate of 66,700 bats made 16 
June 1957 by Constantine (1967) compares poorly with an estimate of 
slightly more than 8,700,000 estimated by Allison (1937) on 16 June 1936. 
Since 1955, National Park Service records report several short periods of bat 
mortality lasting a few days and involving a few to an estimated several 
thousand bats (Constantine 1967; National Park Service Personnel, pers. 
comm.). During a visit by one of the authors ( Altenbach) in September 197 1 , 
several thousand infant bats were found dead on the floor beneath the roost 
area. The condition of the bats suggested they had all died that summer and 
most were less than a few days old at the time of death. 

Concern about the decline in the summer Tadarida population prompted 
the National Park Service to support a detailed study of the bat population 
at Carlsbad Cavern. Initiated in the spring of 1973, this study was to 
determine (1) the extent of the population decline; (2) factors contributing to 
the decline; and (3) the general dynamics of the population each season for 
several seasons. An accurate estimation of the size of the Tadarida popula- 
tion is essential for each of these objectives. 

Humphrey (1971) pointed out that techniques of evening flight counts 
which have provided accurate estimates of population size in small bat 
colonies (Moffat 1905; Venables 1943; Greenhall and Stell 1960; Sluiterand 
VanHeerdt 1966; Baker and Ward 1967; Stebbings 1968; Humphrey and 

341 



342 ALTENBACH ET AL. 

Cope 1968; Watkins 1970) are of little value in estimating size of large 
populations of Tadarida. Humphrey (1971) also pointed out that popula- 
tion estimates of Tadarida by visual inspection of roosting clusters, evening 
flight, or both (Davis et al. 1962; Perry 1965; Constantine 1967; Cockrum 
1969; Twente 1956) "are probably in the correct order of magnitude but 
otherwise of questionable accuracy." 

In the summer of 1973 the entire Tadarida population at Carlsbad Cavern 
roosted on the ceiling of an elongated dome near the east end of the portion 
of the cavern known as Bat Cave. We considered a careful reapplication of 
the technique of multiplying the number of bats per unit area of roost sur- 
face by the total area of occupied roost surface that Constantine (1967) had 
applied at this roost several years before and that other investigators had 
applied elsewhere. However, the ceiling of the roost, which is highly irregu- 
lar and averages roughly 80 ft above the floor, is inaccessible. From below, 
the exact area of the roost cannot be calculated. Rock-climbing techniques 
that could provide access to the roost area are not practical because of the 
decomposed nature of the limestone and the adherence of large mats of 
damp guano to the rock surface. We also agreed that the disturbance of the 
bat colony by an application of this technique was not justified even if it were 
feasible. However, the shape and relative size of the cluster of roosting bats 
could be noted quickly and was useful in determining movements of the bats 
in the roosting area and large changes in population size. 

We also considered a capture-recapture method that had been used at 
Carlsbad Cavern by Constantine (1967) and at Eagle Creek Cave, Greenlee 
County, Arizona, by Cockrum (1969, pers. comm.). This technique 
employed a bat trap, developed by Constantine (1958), which was set in the 
cavern entrance and collected bats during the outflight. Bats caught on one 
night were marked by banding and released. The ratio of marked bats to 
unmarked bats caught on a subsequent night was used to estimate the total 
bat population. We chose not to use this technique because we could not 
make the necessary assumption that a bat captured and marked had an equal 
chance of being recaptured. We also felt that the placement of a bat trap in 
the cavern entrance and resultant disturbance of large numbers of bats 
would not be justified. 

Allison (1937) estimated the size of the Carlsbad population of Tadarida 
on 1 6 June 1 936 by observing the column of bats as they flew from the mouth 
of the cavern. He estimated that during the "full force" of the outflight the 
cross-sectional area of the column was 3 1 4. 1 6 ft 2 (20 ft in diameter), that each 
cubic foot of the column contained one bat, and that the airspeed of the bats 
was 29 ft/ sec. Thus, 29 linear ft passed his observation point each second. 
The evening of his observation the "full force" of the outflight lasted from 
7:03 to 7:17 p.m. From 7: 17 to 7:21 p.m. he estimated bats left the cavern at 
one-half the "full force" rate. By multiplying the bats per second value by 
total seconds for each time period and adding the totals he estimated 



TAD ARID A POPULATION 343 

8,741,760 Tadarida left the cavern that evening. He did not report whether 
any bats remained in the roost after the outflight. 

A more conservative estimate of 3 million had been made by Bailey (1928) 
but in personal communication with Allison (1937) Bailey accepted the 8 
million figure as being more nearly correct. Again by personal communica- 
tion with Allison (1937), Bailey suggested that an accurate estimation of the 
population size would require still and motion picture photography of the 
outflight. 

The first report of the application of a photographic Tadarida population 
estimation technique was that by Humphrey (1971) who took one photo- 
graph every 60 sec of the outflight column from Vickery I Cave, Major 
County, Oklahoma. The bats were photographed while in a confined 
column shortly after leaving the cave and before dispersing to feed. The 
number of bats in each photograph, the speed of the column, and the dura- 
tion of the outflight were used to estimate the total number of bats that left 
the cave each evening. 

The application of a photographic technique at Carlsbad Cavern in 1973 
was more complicated because the column of bats was very erratic once it 
passed the cave mouth, and was of such low density, in contrast to the 
description of Allison (1937), that scattered bats covered an area far too 
large for photographs. Careful observations of the outflight for a period of 
several weeks revealed that the only place the column of bats could be photo- 
graphed was against the ceiling of the first undercut in the mouth of the cave. 
All the bats left the cave through a restricted space beneath this portion of 
the ceiling. 

METHODS 

On the evening of 1 September 1973, a 35-mm camera with a 50 watt- 
second electronic flash and a Miliken 16-mm high-speed motion picture 
camera with a 650-watt movie light were set up on the trail 45 ft beneath the 
critical area of cave ceiling and focused on a point about 3 ft beneath the 
ceiling. Still photographs were taken at 30-sec intervals during the first 44.5 
min of the outflight and at 1-min intervals thereafter (Table 1). Illumination 
was provided by the electronic flash unit and the film was exposed at an ASA 
rating of 1000. Five second long high-speed motion picture runs at 200 
frames per sec were taken at 5-min intervals during the first 45 min of the out- 
flight, using Kodak Four-X Reversal Film. 

Glossy, 8 x 10 in. prints were made from the 35-mm negatives, and bats 
were counted and their direction of flight recorded. A line parallel to ( 7-axis 
line) and another perpendicular to (X-axis line) a line drawn across the 
mouth of the cave were superimposed on each photograph (Fig. 1). 

The motion pictures of the outflight showed that there was a small area of 
space, hereafter called the "exit space," immediately inside the lip of the 
undercut through which all the bats left the cave (Fig. 1). However, only bats 



344 



ALTENBACH ET AL. 



TABLE 1. Data from the photographic population size estimation of Tadarida brasiliensis at 
Carlsbad Cavern. 



Nr 



T„ 



N x T n 



7:24 


10 


7:24:30 


5 


7:25 


1 


7:25:30 


34 


7:26 


27 


7:26:30 


16 


7:27 


17 


7:27:30 


14 


7:28 


19 


7:28:30 


45 


7:29 


47 


7:29:30 


47 


7:30 


16 


7:30:30 


59 


7:31 


6 


7:31:30 


32 


7:32 


17 


7:32:30 


39 


7:33 


35 


7:33:30 


33 


7:34 


65 


7:34:30 


24 


7:35 


30 


7:35:30 


31 


7:36 


38 


7:36:30 


46 


7:37 


21 


7:37:30 


25 


7:38 


42 


7:38:30 


21 


7:39 


30 


7:39:30 


13 


7:40 


11 


7:40:30 


5 


7:41 


12 


7:41:30 


19 


7:42 


20 


7:42:30 


23 



0.35 



0.27 



0.30 



-0.31 



7:43 


26 


7:43:30 


22 


7:44 


20 


7:44:30 


20 


7:45 


27 


7:45:30 


7 


7:46 


25 


7:46:30 


12 


7:47 


22 


7:47:30 


21 


7:48 


24 


7:48:30 


22 


7:49 


28 


7:49:30 


6 


7:50 


26 


7:50:30 


17 


7:51 


17 


7:51:30 


27 


7:52 


19 


7:52:30 


22 


7:53 


21 


7:53:30 


24 


7:54 


13 


7:54:30 


17 


7:55 


17 


7:55:30 


16 


7:56 


13 


7:56:30 


14 


7:57 


18 


7:57:30 


19 


7:58 


20 


7:58:30 


20 


7:59 


12 


7:59:30 


33 


8:00 


9 


8:00:30 


13 


8:01 


9 


8:01:30 


12 



0.30 



■0.29 



•0.29 



■0.30 



10 



8:02 


9 


8:02:30 


9 


8:03 


32 


8:03:30 


18 


8:04 


14 


8:04:30 


27 


8:05 


13 


8:05:30 


16 


8:06 


17 


8:06:30 


17 


8:07 


7 


8:07:30 


18 


8:08 


16 


8:08:30 


2 


8:09 


11 


8:10 


11 


8:11 


16 


8:12 


22 


8:13 


I 


8:14 


9 


8:15 


10 


8:16 


4 


8:17 





8:18 


6 


8:19 


5 


8:20 


5 


8:21 


4 


8:22 


3 


8:23 





8:24 


1 


8:25 






•0.30 



■0.27 



x = time of still photograph (MDST); 
N x - number of bats leaving cave in photograph at time x; 
n - motion picture run number; 

T n - replacement time (in seconds) used as an estimator within time period enclosed in 
brackets. 



oriented with the long-body axis inclined outward from a Y-slxis line in a 
photograph (Fig. 1 ) seemed committed to exit from the cave. Bats with long- 
body axis parallel to or inclined inward from a 7-axis line (Fig. 1) seemed 



TAD ARID A POPULATION 345 




Fig. 1. Photograph of the ceiling of the entrance area of Carlsbad Cavern, showing 
areas used in the photographic population estimation technique. 



committed to fly back into the cave mouth. The motion picture data showed 
that less than 0. 1 percent of the bats in the exit space failed to behave in this 
manner. The motion pictures also showed that a few bats, which had 
apparently left the cave, flew back into the cave through the exit space. This 
number of in-flying bats was less than 1 percent of the number of out-flying 
bats and was accounted for by subtracting the number of bats in the exit 
space with their long-body axis oriented directly into the cave from the num- 
ber of out-flying bats. Thus a number of out-flying bats in a still photograph 
taken at instant X was computed (N x ). 

Bats moving through the exit space parallel to the X-axis line (Fig. 1) 
crossed the space in 0.20 sec. Those moving through the exit space at greater 
angles from the X-axis line or in arcs took more time to cross the space. By 
projecting the films repeatedly and timing the flight of individual bats across 
the exit space, an average time period ( T n ) for bats to cross the exit space 
was computed for each motion-picture sequence. This time period was con- 
sidered the time (replacement time) required for the out-flying bats photo- 
graphed in the exit space at one instant to be replaced by a new group of bats. 
Each T n value was used as the replacement time for out-flying bats over a 
time period for which T n is the estimator (Table 1). For example, during 
each of the time periods before and after the motion picture run at 7:29 p.m. 



346 ALTENBACH ET AL. 

(the periods with the instants 7:27-7:31:30 as their midpoints), a group of 
bats in the exit space was assumed to be replaced by another group of bats 
every 0.27 sec (Table 1). 

Each photograph was used to estimate the number of bats leaving the cave 
during a time period (P x ) which had the instant of the photograph as its mid- 
point. The photograph of 7:24, which is an estimator of only the 15-sec 
period preceding it and the 30-sec period following it, and the photograph of 
8:09, which is an estimator of the 15-sec period preceding it and the 30-sec 
period following it, are the only exceptions (Table 1). A T n value was divided 
into each of the time periods (Px values for which it was an estimator, to 
calculate the number of replacements of out-flying bats, R x ) in the exit space 
during each of those time periods. Each R x value was multiplied by the 
appropriate N x value to produce an estimate of the number of bats that left 
the cavern during each time period. This procedure is summarized as 
follows: P x IT n = R x and R X (N X ) = E x \ 
where 

P x - time period for which photograph at time X is an estimator; 

T n = average time for bats to cross exit space during the time period for 
which motion-picture run is an estimator (replacement time); 

R x - number of replacements of bats in the exit space during time period 

p x ; 

N x - number of bats leaving the cavern in photograph at time X\ 

E x - estimate of the total number of bats leaving the cavern during the 

time period P x . 
Table 1 shows the time each still photograph was taken (x), the N x value 
for each still photograph, the time of each motion-picture run («), and the 
time period for which it is an estimator. By adding each of the E x values, we 
produced an estimate of the total bats that left the cavern the evening of 1 
September 1973. 

RESULTS AND DISCUSSION 

The total number of bats that left the cave the evening of 1 September 1973 
was calculated to be 218,153. No bats were seen in the roost area immedi- 
ately after the outflight. Tadarida born at Carlsbad Cavern in 1973 began 
flying an average of 39 days after birth. This period is similar to those 
reported by Davis et al. (1962) for Texas Tadarida colonies and by Constan- 
tine (1967) for Carlsbad Cavern during the 1957 season. The first outflight 
sample in which all of the nonbarren females were postpartum was that of 14 
July (Fig. 2). In the sample of 6 July, 95% of the nonbarren females were 
postpartum. Thus, by 14 August, 95% of the surviving young of the year 
should have been flying, and by 22 August, 9 days before our population 
estimation, all of the 1973 young should have been flying (Fig. 2). The size 
and position of the cluster of roosting bats were carefully observed over 
the period of 12 days preceding 1 September. We could detect no changes 



TA DA RID A POPULATION 



347 



percent non barren ¥$ postpartum 



39 days 



percent young inoutflight 



16 22 29 

JUNE 



6 14 19 26 

JULY 



2 9 15 21 

AUGUST 



Fig. 2. Data from bats captured during the outflights from Carlsbad Cavern during 
June, July, August, and September 1973. The graph on the left is the percent of non- 
barren females in the population which were postpartum. 



in the relative size of the cluster during this time. The duration and the den- 
sity of the outflight also remained about the same during this period. Thus 
we assumed no immigration or emigration during the 12-day period pre- 
ceding our population estimation procedure. We feel our estimate is more 
accurate than any made heretofore and represents a relatively stable popu- 
lation of adult males, barren females, nonbarren postpartum females, and 
all surviving young of the year. Data on sex and age composition of the 
population during, before, and after 1 September 1973 have been 
assembled and are the subject of another report. 

This technique for population size estimation was not used again during 
1973. However, it was used during the summer of 1974 and will be used at 
monthly intervals during the summer of 1975 and 1976. We feel confident 
that we have a useful and valid means of estimating the size of large Tadarida 
populations with minimal disturbance of the bats. This technique should be 
applicable at any Tadarida colony and facilitate additional population 
studies on this bat. 



LITERATURE CITED 

Allison, V. C. 1937. Evening bat flight from Carlsbad Caverns. J. Mammal. 
18:80-82. 

Bailey, V. 1928. Animal Life of the Carlsbad Cavern. Am. Soc. Mammal. 
Monogr. 3, 195 p. 

Baker, R. J., and C. M. Ward. 1967. Distribution of bats in southwestern Arkan- 
sas. J. Mammal. 48:130-132. 

Cockrum, E. L. 1969. Migration of the guano bat, Tadarida brasiliensis. Pages 



348 ALTENBACH ET AL. 

303-336 in J. Knox Jones, Jr., ed. Contributions in Mammalogy. Misc. Publ. 

Mus. Nat. Hist. Univ. Kans. 51:1-428. 
Const antine, D. G. 1958. An automatic bat-collecting device. /. Wildi. Manage. 

22:17-22. 
1967. Activity patterns of the Mexican free-tailed bat. Univ. N. M. Bull. 

Biol. Ser. 7:1-79. 
Davis, R. B., C. F. Herreid, II, and H. L. S hort. 1962. Mexican free-tailed bats 

in Texas. Ecol. Monogr. 32:311-346. 
Greenhall, A. M., and G. Stell. 1960. Bionomics and chemical control of free- 
tailed house bats (Molossus) in Trinidad. U.S. Fish Wildl. Serv. Spec. Sci. Rep. 

Wildi 53:1-20. 
Humphrey, S. R. 1971. Photographic estimation of population size of the Mex- 
ican free-tailed bat, Tadarida brasiliensis. Am. Midi. Nat. 86(l):220-223. 
and J. B. Cope. 1968. Records of migration in the evening bat, Nycticeius 

humeralis. J. Mammal. 49:329. 
Moffat, C. B. 1905. The duration of flight among bats. Irish Nat. 14:97-109. 
Perry, A. E. 1965. Population analysis of the guano bat Tadarida brasiliensis 

(Saussure) using lens-weight method of age determination. Ph.D. Thesis, 

Oklahoma State Univ., Stillwater, 60 p. 
Sluiter, J. W., and P. F. VanHeerdt. 1966. Seasonal habits of the noctule bat 

(Nyctalus noctula). Arch. Ne'er. Zool. 16:423-439. 
Stebbings, R. E. 1968. Movements, composition and behaviour of a large colony 

of the bat, Pipistrellus pipistrellus. J. Zool. London 156:15-33. 
Twente, J. W., Jr. 1956. Ecological observations on a colony of Tadarida mex- 

icana. J. Mammal. 37:42-47. 
V enables, L. S. V. 1943. Observations at a pipistrelle bat roost. J. Anim. Ecol. 

12:19-26. 
W atkins, L. C. 1970. Observations on the distribution and natural history of the 

evening bat {Nycticeius humeralis) in northwestern Missouri and adjacent Iowa. 

Trans. Kans. Acad. Sci. 72:330-336. 



ACKNOWLEDGMENTS 

The cooperation and assistance of the Carlsbad Caverns National Park 
Service personnel are greatly appreciated. We especially thank Philip Van 
Cleave and Charles Peterson. We also wish to thank Scott Berger for his 
assistance in analyzing the still photographs and motion pictures and James 
Findley for critically reviewing this manuscript. Financial support was 
obtained from the National Park Service, The World Wildlife Fund, and the 
U.S. Fish and Wildlife Service. 



Coexistence of Two Species of 
Kangaroo Rats (Genus Dipodomys^ 
in the Guadalupe Mountains 
National Park, Texas 



MARGARET A. O'CONNELL, Texas Tech University, 
Lubbock 

Several species of heteromyid rodents are often found in coexistence in the 
deserts of North America (Hawbecker 1951; Reynolds 1958; Chew and 
Butterworth 1964; Brown 1973; Rosenzweig 1973; Reichman 1975). Six 
species of heteromyid rodents, Dipodomys ordii, D. merriami, D. 
spectabilis, Perognathus penicillatus, P. flavus, and P. intermedius, have 
been found in the western portion of the Guadalupe Mountains National 
Park, Texas. These rodents have broadly similar ecological needs. They are 
basically granivorous, nocturnal, and burrowing. Whenever species with 
similar ecological needs are found in sympatry, the possibility of inter- 
specific competition arises. The widespread coexistence of different species 
of heteromyid rodents suggests that these animals have evolved with mech- 
anisms which reduce competitive elimination. Models that explain the co- 
existence of similar and dissimilar sympatric species by habitat selection and 
resource allocation have been constructed by Mac Arthur and Levins (1964). 
Several workers have applied these models to coexistence in heteromyid 
rodent populations. Rosenzweig (1973) has shown habitat selection to be an 
important factor in the coexistence of Dipodomys and Perognathus. Seed 
size has been proposed as a mechanism of resource allocation among hetero- 
myids (Brown and Lieberman 1973). Recently, Reichman (1975) reported 
that Dipodomys and Perognathus specialize on different proportions of the 
same food resources. 

A prerequisite to understanding allocation of food resources in areas of 
sympatry is a knowledge of the animals' diets. The primary purpose of this 
rtudy was to determine the diets of two species of kangaroo rats, Dipodomys 
nerriami and D. ordii, through microscopic analysis of stomach contents, to 
examine dietary overlap in various seasons and habitats, and to evaluate 
hese data in light of models for coexistence. These two species of 

349 



350 O'CONNELL 

Dipodomys were chosen for study because they are relatively abundant in 
the area and, although easy to distinguish in the field, they are of similar size 
and morphology. The food habits of heteromyids have generally been deter- 
mined by examination of cheek-pouch contents (Reynolds 1958; Dunham 
1968; Gaby 1972; Chapman 1972). These workers have assumed that 
cheek-pouch contents are an accurate indicator of the actual diet of the 
rodent. Thus, a second objective of this study was to compare cheek-pouch 
contents and stomach contents in these two rodents. 

The western portion of the Guadalupe Mountains National Park was not 
included within the proposed Wilderness Area of the 1974 Master Plan for 
the park (U.S. Department of Interior 1974). According to this document, 
this area may be transversed by roads. Thus it is presently in a state of transi- 
tion; in the past it has been heavily grazed, in the future it may face human 
impact of a different nature. Because interspecific relationships are often of a 
very tenuous nature, any alteration in habitat may favor one species over the 
other. A further purpose of this study was to provide the baseline data from 
which these two kangaroo rats may be used as biological indicators of the 
changing range conditions and the degree of human impact in this area of the 
national park. 

DESCRIPTION OF THE STUDY AREA 

The study is located in the western portion of the Guadalupe 
Mountains National Park, Texas. It is characterized by bajadas extending 
westward from the sheer limestone escarpment that forms the western edge 
of the Guadalupe Mountains. Elevation ranges from 1430 m at the base of 
the escarpment to 1 1 20 m on the western boundary. 

This area includes a combination of features from the grassland and desert 
scrub formations (Warnock undated; Burgess and Northington 1977). The 
western boundary of the park is in close proximity to an extensive salt flat. 
Atriplex canescens (four-wing-saltbush) and Sporo bolus airoides (alkali 
sacaton) are the dominants on these saline soils. Beyond these soils, Atriplex 
gives way abruptly to Larrea tridentata (creosote bush) which is the major 
dominant of the area. On the lower portions of the bajada, Larrea is found in 
association with Prosopis glandulosa var. torreyana (honey mesquite), 
whereas nearer the escarpment its associates include Bouteloua eriopoda 
(black grama) and Erioneuron pulchellum (fluffgrass). Throughout the area 
various species of Yucca (Spanish bayonet) and Opuntia (prickly pear) are 
present. 

The temperatures for the area ranged from an average low of -2.4° C in 
December 1974 to an average high of 27.2°C in June 1974 (U.S. Depart- 
ment of Commerce 1974). Precipitation was relatively low, 137.9 mm during 
the 10 months prior to the the study. The first 5 months of the study 
(February through June 1974) were also dry, with 17.2 mm of precipitation. 
The area received 329.6 mm of rain from July 1974 through January 1975, 
and 102.4 mm of snow in December 1974 and January 1975. 



COEXISTENCE OF DIPODGMYS 



351 




Fig. 1. Distribution of the two species of Dipodomys in the western portion of the 
Guadalupe Mountains National Park, Texas. The numbers denote the five main 
trapping localities of this study. 



352 O'CONNELL 

METHODS AND MATERIALS 

The populations were sampled a minimum of once a month from 
February 1974 through January 1975, with a total of 5728 trap-nights. 
Museum Special snap traps and Victor rat traps were placed at 4-m inter- 
vals along a transect. 

Initially, trapping was conducted to determine the distribution of 
Dipodomys merriami and D. ordii within the western portion of the park. 
Once distribution patterns were established, trapping was limited to five 
main areas (Fig. 1). 

Rodents were either prepared in the field or frozen on dry ice and pre- 
pared in the laboratory. Stomachs were removed, their contents air dried, 
and cheek-pouch contents were collected. Stomach contents of 276 D. 
merriami and 58 D. ordii were examined for dietary analysis. Even though 
population data were not collected, field observations suggest that these 
numbers probably reflect the relative densities of the two species in the area. 

Stomach contents were prepared using the procedures outlined by 
Baumgartner and Martin (1939), Dusi (1949), Hansen and Flinders (1969), 
and Hansson (1970). Contents were washed over a 200-mesh screen sieve and 
thoroughly mixed. Randomly selected aliquots of the contents of each 
stomach were placed on two microscope slides and cleared with several 
drops of Hertwig's solution (Baumgartner and Martin 1939). The Hertwig's 
solution was evaporated by boiling, and the material was spread evenly over 
the slide. H oyer's solution was added as the mounting medium and the slides 
were dried at 55° C for 48 hours. 

A reference collection was made of the plants in the area. Plants were 
collected, dried, and ground, using either a Waring blender or a Wiley Mill. 
Reference material was prepared on microscope slides in the same manner 
as the stomach contents, with the exception that Permount was used as the 
mounting medium. 

Ten nonoverlapping microscope fields (100*) were examined from each 
of the two slides made from the contents of each rodent's stomach. The 
presence of each food item in each microscope field was recorded. Percent 
frequency for each food item was determined as the number of times it 
appeared in a field expressed as a percentage of the total number of fields 
examined (Flinders and Hansen 1972). 

Percent frequency can be converted to particle density per field (Fracker 
and Brischle 1940), and relative density can then be used to estimate percent 
dry weight of food items in diets (Sparks and Malechek 1968). Sparks and 
Malechek (1968) based their estimation of percent dry weight on samples 
containing known amounts of grasses and forbs that had been artificially 
mixed prior to analysis. Many workers in diet studies have successfully used 
this procedure (Ueckert and Hansen 1971; Flinders and Hansen 1972); how- 
ever, these studies have dealt primarily with animals whose diets were 
basically green vegetation, i.e., leaf and stem material. Unlike green vegeta- 
tion, different seed types vary considerably in size and specific weights. 



COEXISTENCE OF D/PODOMYS 353 

Because kangaroo rats are primarily granivorous, this procedure was not 
used in this study, and the data are based on percent frequency. 

Kulzynski's Similarity Index (from Oosting 1956) was used to indicate the 
similarity of diets between interspecific and intraspecific samples as to sex, 
time, and locality. This expression is calculated by the formula: 



2 StoMlOOl/Sfli+h 



where a t represents the mean percentage of food item i in the diet of group X, 
bi represents the mean percentage of food item i in group Y, and w, 
represents a t if a, < bi and b { if bi < a*. 

Cheek-pouch contents were used as a separate indicator of these animals' 
diet. The dry-weight composition of the cheek pouches was determined as 
the weight of one type of food item per total weight of all food items. Seeds 
found in the cheek pouches were identified by comparison with a reference 
collection of seeds from the study area. Cheek-pouch contents of 192 D. 
merriami and 9 D. ordii were examined. 

The number of food items appearing in each animal's stomach and cheek 
pouch was recorded. The mean number of food items per stomach and per 
cheek pouch was determined for each species during the different months of 
the year. The BMD, 07 V computer program was used to perform an analysis 
of variance on these data and to separate the means into homogeneous sub- 
sets using the Student-Newman-Keuls test (Sokal and Rohlf 1969). 

RESULTS 
Distribution 

The trapping revealed an almost continual distribution of Dipodomys 
merriami throughout the entire study area and a scattered distribution of D. 
ordii (Fig. 1). The only location from which D. merriami was not collected 
was the white gypsiferous sand dunes on the western boundary of the park. 
Dipodomys ordii was collected at the lower elevations near the southern 
entrance to the park (1230 to 1340 m), the area north of the Patterson Hills 
(1220 m), and the westernmost portion of the park (1230 to 1350 m). For a 
complete account of all specimens collected and locality descriptions, see 
Genoways et. al. (1977). 

During the study, Dipodomys spectabilis was collected for the first time 
from the Guadalupe Mountains National Park. Although trapping success 
for this larger kangaroo rat has been low, field observations of active 
burrows suggest that the distribution of D. spectabilis is also continuous 
throughout the western portion of the park. 

In addition to kangaroo rats, three other heteromyid rodents, 
Perognathus penicillatus, P. intermedius, and P.flavus, have been collected 
from this area. These smaller heteromyids also have a continuous distribu- 



354 



O'CONNELL 



tion in this area. The diets of these pocket mice are being analyzed and will be 
reported at a later date. 



50 



40 

Z 

o 

" 30 
O 

a. 

O 

w 20 






10 




D. M ERR I AM I 
D. OR DM 



Z 

V) 



O 

70 



70 




o 

70 
00 



c 

c\ 

c 



z 

-I 
en 



Fig. 2. Mean percent composition based on relative frequency of major food items in 
the annual diets of Dipodomys merriami and D. ordii as determined from stomach 
analyses. Greenery refers to leaf and stem material. Grasses, shrubs, forbs, and succu- 
lents refer to seeds of these plant types. 



COEXISTENCE OF DIPODOMYS 



355 



Diet 

Seeds were the most important food item for both Dipodomys merriami 
and D. ordii. In addition, insects and green vegetation were important in the 
diets of both species. 

Seeds made up 63.6% of the diet of D. merriami, with shrub seeds con- 
stituting 23%, forb seeds, 24.1%, grass seeds, 4.5%, and succulent plant 
seeds, 12% of the total diet (Fig. 2). The diet of D. merriami varied 
seasonally (Fig. 3). Seeds, insects, and green vegetation were present in the 
diet throughout the year; however, during the late summer and autumn, 



30 



20 



!0 



C >1 



D. merriami 




SEEDS 



> 2 



c c 

Z i- 



> en 

c m 

O -i 

C m 

(/> 2 

-• 00 



m > 
m Z 

2 > 



Fig. 3. Seasonal variation of the three main food items in the diet of Dipodomys 
merriami, as determined from microscopic examination of stomach contents. 



356 



O'CONNELL 



both insects and green vegetation were less important. Green vegetation 
was most important during mid-summer. Insects were eaten in greatest 
quantities during the winter months. The seasonal variation of the four 
seed types is shown in Fig. 4. The high percentage of succulent plant seeds 
during April and May reflects the importance of Dasylirion (sotol) seeds in 
the diet at this time. Shrub seeds became more important in the diet during 
mid-summer, corresponding to the seed-setting of both Prosopis and 
Larrea. During the autumn, forb seeds became very important, with 
Euphorbia seeds the major component of the diet. 



10 



D. merriami 




"" 2 

00 > 

I* 

3D 

-< 



> 2 

■o > 



c c 

Z i- 



> m 

C m 

<* 2 



O Z 

Qo 

O m 

03 J 

m ^ 



m > 

m c 

2 > 

03 TO 



Fig. 4. Seasonal variation of the four major seed types in the diet of Dipodomys 
merriami, as determined from microscopic analysis of stomach contents. 



COEXISTENCE OF DIPODOMYS 



357 



Of the 33 food items identified in the diet of D. merriami, 16 items made up 
over 1% (Table 1). The most important food items included Larrea, insects, 
Euphorbia, green vegetation, Opuntia, Prosopis, and Lepidium (pepper- 
grass). 

Seeds were more important in the diet of D. ordii, making up 89.5% of the 
total diet. Shrub seeds constituted 15.5%, forb seeds, 27%, grass seeds, 41%, 
and succulent plant seeds 6% (Fig 2). The seasonal variation of the three 
main food types is shown in Fig. 5. Seeds were the major component 



TABLE 1 . Percent composition of food items in the mean annual diets of Dipodomys merriami 
and D. ordii as determined from stomach analyses. 



Food item 



Common name 



D. merriami 



D. ordii 



Larrea 


Creosote bush 


Insects 




Prosopis 


Mesquite 


Greenery 




Bouteloua 


Grama grass 


Lesquerella 


Bladder-pod 


Lepidium 


Peppergrass 


Yucca 


Spanish bayonet 


Flourensia 


Tarbush 


Sporobolus 


Dropseed 


Atriplex 


Saltbush 


Euphorbia 


Spurge 


Croton 


Rosval 


Opuntia 


Prickly pear 


Gutierrezia 


Snakeweed 


Dyssodia 


Dogweed 


Chilopsis 


Desert willow 


Nerisyrenia 




Dasylirion 


Sotol 


Pedis 




Tridens 


Slim Tridens 


Oenothera 


Evening primrose 


Kallstroemia 




Erioneuron 


Fluff grass 


Sphaeralcea 


Globe mallow 


Fallugia 


Apache-plume 


Erodium 


Alfilerillo 


Tetraclea 




Coldenia 




Poliominthia 


Rosemary mint 


Bahia 




Tides tomia 




Krameria 


Range ratany 


Fouquieria 


Ocotillo 



14.2 


5.1 


23.5 


3.1 


4.2 


4.1 


12.9 


7.5 


3.1 


17.8 


Tr 


1.1 


4.6 


Tr 


3.7 


1.7 


1.7 


Tr 


1.4 


17.0 


2.4 


6.4 


10.2 


2.0 


Tr 


4.6 


4.0 


3.0 


1.5 


1.0 


2.2 


Tr 


Tr 


1.6 


Tr 


8.8 


4.1 


Tr 


1.5 


Tr 


Tr 


1.1 


Tr 


3.1 


Tr 


2.4 


Tr 


1.4 


Tr 


1.1 


Tr 


1.9 


Tr 


Tr 


Tr 


Tr 


Tr 


Tr 


A 


Tr 


Tr 


Tr 


Tr 


Tr 


Tr 


Tr 


Tr 


Tr 



Tr = a trace, or less than 1%; A = absence from the diet. 



358 O'CONNELL 





D. o r d i i 






100 
















GRE ENE RY 








90 
















V ^•^* insecT^jC^ 








80 












70 












"V 












m 












TO 












r» 60 




SEEDS 








m 












z 












- 50 












n 












O 












2 40 












"O 












O 












1 30 












-« 












° 20 












10 












■" 5 


> > «_ «_ > t/» 


o z 


O «- 


lARC 
EBRU 


E PTE 
UGU 

ULY 
UNE 

lAY 
PRI L 


n O 

-• < 

O ni 


ANU 
ECEN 


> I 


S 2 


" i 


03 > 


90 
-< 


03 

m 
90 


03 
50 m 

70 


rn so 
» -< 



Fig. 5. Seasonal variation of the three main food items in the diet of Dipodomys 
ordii, as determined from microscopic analysis of stomach contents. 



throughout the year, whereas green vegetation was most important during 
the spring and late summer. Insects were relatively constant throughout the 
year, becoming less important during the late summer and, unlike D. 
merriami, during the winter. 

The seasonal variation of the four seed types is shown in Fig. 6. Succulent 
plant seeds were present in the diet with relatively constant precentages 
throughout the year. The increased importance of forb seeds during the 



COEXISTENCE OF DIPODOMYS 



359 



100 
90 
80 




Z 

H 50 

r» 

O 

2 40 

o 



</> 



30 



D. o r d i i 



FORBS 



GRASSES 




SHRUBS 



m * 

OB > 

» » 

c « 

> I 

3D 



> 2 



c c 

1 < 



> v» 

c m 

^ m 

3 5 



° 5 

n O 

■H < 

O m 



m ^ 

n Z 

cp > 

m » 

» -< 



Fig. 6. Seasonal variation of the four major seed types in the diet of Dipodomys 
ordii, as determined from microscopic analysis of stomach contents. 



spring reflects the amount of Lepidium eaten at this time. As in the diet of D. 
merriami, shrub seeds became more important during mid-summer. Grass 
seeds were eaten in greater quantities during the late summer and autumn. 
Twenty of the 34 food items identified in the diet of D. ordii contributed 
over 1% of the diet (Table 1). Larrea, Bouteloua, green vegetation, 
Sporobolus, Atriplex, and Nerisyrenia were the major components of the 
diet. 



360 



O'CONNELL 



Comparison of Cheek-Pouch and Stomach Contents 

If cheek-pouch contents rather than stomach contents are used to 
describe the diet, the picture is considerably different. Seeds made up 99.2% 
of the diet of D. merriami, based upon cheek-pouch contents. Only one 
insect was found in the cheek pouches of the 276 specimens examined 
(Fig. 7). 

The most important seeds found in the cheek pouches were Larrea, 
Prosopis, Opuntia, and Euphorbia. Although these seeds were also 
important in the stomach contents, they appeared in different proportions in 
the cheek pouches (Fig. 7). For example, Larrea seeds constituted 14.2% of 
the diet as determined from stomach contents, but made up 37.2% of the diet 
as determined from cheek-pouch contents. Some seeds, such as Lepidium, 
were relatively important in the stomach contents but insignificant in the 
cheek-pouch contents (Fig. 7). Conversely, other seeds, such as Chilopsis, 



50 



40 



* 30 
o 

Ol 

o 

w 20 




STOMACH 



10 



n 



Fig. 7. Comparison of the contribution of 10 food items in the diet of Dipodomys 
merriami as determined by examination of cheek-pouch contents to the contribu- 
tion of these foods as determined by examination of stomach contents. Percent 
composition was determined on a dry weight basis in cheek-pouch samples and on a 
relative frequency basis in stomach samples. 



COEXISTENCE OF DIPODOMYS 361 

appeared in the stomach contents with a percentage less than one, whereas 
they constituted 6.8% of the diet as determined by cheek-pouch contents. 
The importance of four food items including insects, greenery, Lepidium, 
and Dasylirion were grossly underestimated by cheek-pouch examination, 
whereas six foods including Larrea, Prosopis, Opuntia, Chilopsis, 
Krameria, and Setaria were greatly overestimated by cheek-pouch 
examination. 

These trends also appeared in the comparison of cheek-pouch contents 
with stomach contents of D. ordii', however, only 15% (nine) of the cheek 
pouches of D. ordii contained food items. This small sample size was 
considered inadequate to determine percent composition. 

Variation in Number of Food Items in Diets 

The mean numbers of food items per stomach were 4.8 ± 0.8 (standard 
deviation) for Dipodomys merriami and 5.6 + 1.0 for D. ordii for the entire 
year. Analysis of variance revealed significant (P < 0.05) seasonal dif- 
ference between the mean numbers of food items. 



TABLE 2. The mean number of food items appearing per stomach contents of Dipodomys 
ordii during the different seasons. 

Season Mean number of food items 

February 1974 6.4 ± 1.2 

May-June 1974 5.2 + 1.0 

August 1974 4.1 ± 1.0 

January 1975 6.7+1.2 



For D. ordii, months represented with a sufficient sample size were 
separated into two homogeneous subsets: — (1) February 1974 and January 
1975, and (2) May-June and August 1974. The mean number of food items 
per stomach for the first subset was 6.1 ± 0.6 and was 4.5 ± 0.6 for the 
second subset (Table 2). 

The means number of items per D. merriami stomachs varied seasonally 
(Table 3). February, March, April, December 1974, and January 1975 fell 
into one homogeneous subset with a mean number of items of 5.5 ± 0.7. A 
second subset included May, June, August, September, October, and 
November 1974 with a mean of 4.0 ± 0.6 food items per stomach. 

The overall average number of food items per cheek-pouch was 1.4 ±0.2 
for D. merriami (Table 3). The mean number of food items in cheek pouches 
did not vary significantly during the different months of the year (P< 0.05). 



362 O'CONNELL 



TABLE 3. The mean number of food items appearing per stomach and cheek pouch contents 
of Dipodomys merriami during the different seasons. 



Mean number of food items 



Season 



Stomach contents 


Cheek-pouch contents 


6.0+ 1.1 


1.6 + 0.7 


5.2+ 1.2 


1.3 ±0.6 


5.2 + 0.9 


1.8 ±0.9 


5.5 ± 1.2 


1.3 ±0.4 


5.1 ± 1.0 


1.7 ±0.8 




1.4 + 0.6 


4.0 ± 0.6 




3.5 ± 0.6 




3.6 ± 0.9 




3.1 ± 1.0 


1.2 ±0.4 


3.9 ± 0.9 


1.5 ±0.5 


4.1 ± 1.0 


1.5 ±0.8 


5.0 ± 1.0 


1.6 ±0.8 




1.2 ±0.5 


5.9 ± 1.2 




7.0 ± 1.2 





February 1974 
March 1974 
April 1974 
May 1974 
June 1974 
August 1974 
Area no. 3 
Area no. 2 
Area no. 1 
September 1974 
October 1974 
November 1974 
December 1974 
January 1975 
Area no. 3 
Area no. 5 



Interspecific Dietary Similarity 

A comparison of the mean annual diets of D. merriami and D. ordii 
yielded a similarity index of 42 (Table 4), which indicates no serious dietary 
overlap. The diets of the two species overlapped on 33 food items (Table 1); 
however, a food consumed in fairly large proportions by one species was 
usually of minor importance in the diet of the other species. 



TABLE 4. Interspecific similarity indices for mean annual diets of Dipodomys merriami and 
D. ordii, and for selected seasonal diets in two habitats. 

Comparison Similarity index 

Total year 42 

Area no. 1 

February 1974 36 

August 1974 55 

Area no. 5 

January 1975 65 



Interspecific dietary similarity varied both with season and location 
(Table 4). The similarity index for the diets of the two species for area no. 1 
(Fig.l), the southern entrance to the park, during February 1974 was 36. 



COEXISTENCE OF DIPODOMYS 363 

This low similarity index reflects the fact that D. merriami consumed more 
insects than did D. ordii. Insects made up 28.7% of the diet of D. merriami 
and only 2.8% of the diet of D. ordii. Bouteloua seeds did not appear in the 
diet of D. merriami at this time, but they made up 34.7% of the diet of D. 
ordii. 

During August 1974 at the same location, the similarity index between the 
diets of the two increased to 55. Insects remained more important in the diet 
of D. merriami, 25.3% as compared to 3. 1% in D. ordii, and Bouteloua seeds 
were still more important to D. ordii (17%) than to D. merriami (1.3%). 
However, Prosopis seeds, which had not been of great importance in either 
kangaroo rat's diet during February, contributed 22.4% and 20.5% to the 
diets of D. merriami and D. ordii, respectively. 

The similarity index for the two diets for a quartz sand-hill habitat (area 
no. 5; Fig. 1) during January 1975 was 65. Both species of rodents had more 
food items per stomach at this locality than at any other (Tables 2 and 3). 
These sand hills had 33 plant species, which is more than three times the 
number found on the surrounding creosote-bush flats (Burgess and 
Northington 1977). 

Intraspecific Dietary Similarity 

The intraspecific comparisons indicate no dietary difference between 
sexes in either species. The index of similarity between males and females 
was 90 for D. merriami (Table 5) and 89 for D. ordii (Table 6). The high simi- 
larity indexes for conspecific males and females indicate that the low inter- 
specific similarities are probably not due to sampling error. 

Intraspecific comparisons also suggest that both kangaroo rats shift their 
diets with respect to time of year and location (Tables 5 and 6). The diet of 
specimens of D. merriami collected from area no. 1 (Fig. 1), the southern 
entrance to the park, was compared for three different times of the year — 
February, August, and November 1974 (Table 5). The similarity index 
between February and August was 48 and that between February and 
November was 40. The diets during August and November showed little 
dietary overlap (similarity index = 28). During February, this kangaroo rat 
feeds on a small percentage of many different food items; however, during 
August and November, D. merriami selects larger quantities of fewer items. 
These selected items are different during August {Prosopis) than during 
November (Euphorbia). 

At the same locality, the similarity index between February and August 
diets of D. ordii was 53 (Table 6). However, only one animal was collected 
from this area during November, making determination of the similarity 
index impossible. 

Another locality (area no. 3; Fig. 1), the area to the north of the Patterson 
Hills, was compared during five different months — May, June, August, 
October 1974, and January 1975. This area is characterized by more open 
creosote flats, with less grass cover than the locality discussed above. The 



364 O'CONNELL 

TABLE 5. Intraspecific similarity indexes for the diet of Dipodomys merriami. 

Comparison Similarity index 

90 



48 
40 
28 

67 
69 
51 
54 
77 
47 
63 
48 
75 
60 



62 

55 

56 
65 
59 
55 
49 
40 
44 
50 

12 



similarity indexes were higher (40 to 77) throughout the year at this locality. 
The spring and summer months were more similar to each other and to a 
winter month than they were to an autumn month. However, the dietary 
overlap betweeen the autumn and winter months was also high. 

The diet of D. merriami throughout various localities on the creo- 
sote-bush flats (areas no. 1, 2, 3, 4; Fig. 1) was similar (similarity index = 55 
to 65) during August. However, when these localities were compared to the 
quartz sand hills (area no. 5; Fig. 1) during the same month, there was con- 
siderably less similarity (similarity index = 40 to 49) (Table 5). When a creo- 
sote-bush flat area (area no. 3; Fig. 1) was compared to the quartz sand hills 
(area nc. 5; Fig. 1) during January 1975, there was even less dietary overlap 
(similarity index = 1 2). There was also very little similarity (similarity index = 



Males vs. females 








Same Localit> 


— Different Seasons 


Area no. 1 








February vs. August 






February vs. November 




August vs. November 




Area no. 3 








May vs. June 








May vs. August 






May vs. October 






May vs. January 






June vs. August 






June vs. October 






June vs. January 






August vs. October 






August vs. January 






October vs. January 






Same season- 


-Different localities 


August 1974 








Area no. 1 vs. 


Area 


no. 


3 


Area no. 1 vs. 


Area 


no. 


2 


August 1974 








Area no. 1 vs. 


Area 


no. 


4 


Area no. 3 vs. 


Area 


no. 


2 


Area no. 3 vs. 


Area 


no. 


4 


Area no. 2 vs. 


Area 


no. 


4 


Area no. 5 vs. 


Area 


no. 


1 


Area no. 5 vs. 


Area 


no. 


2 


Area no. 5 vs. 


Area 


no. 


3 


Area no. 5 vs. 


Area 


no. 


4 


January 1975 








Area no. 5 vs. 


Area 


no. 


3 



COEXISTENCE OF DIPODOMYS 365 

TABLE 6. Intraspecific similarity indexes for the diet of Dipodomys ordii. 

Comparison Similarity index 

Males vs. females 89 

Same locality — Different seasons 
Area no. 1 

February vs. August 53 

Area no. 5 

August vs. January 64 

Same season— Different locality 
August 1974 

Area no. 3 vs. Area no. 1 59 

January 1975 

Area no. 5 vs. Area no. 4 13 



13) at this same time between the diet of D. ordii on the quartz sand hills 
(area no. 5; Fig. 1) and another creosote-bush flat location (area no. 4; Fig. 
1) (Table 6). 

DISCUSSION 
Diet 

The food habits of Dipodomys merriami and D. ordii are broadly similar. 
Both species are primarily granivorous but also eat insects and green vegeta- 
tion. Of the 34 food items found in the diet of D. ordii, 33 also appeared in 
the diet of D. merriami. However, the proportions of these food items in the 
diet vary greatly between species. This finding gives support to the model of 
Mac Arthur and Levins (1964) that animals with similar diets will tend to 
specialize on specific proportions of the food resources. Mac Arthur in a later 
paper (MacArthur and Pianka 1966) suggested that this may occur because 
competitor animals will utilize fewer patches rather than reduce their diets. 
According to these authors, the high number of food items shared by these 
two species would suggest that the animals are foraging in different habi- 
tats, or patches. Rosenzweig (1973) has shown D. merriami to prefer open 
creosote-bush habitats with little cover. Field observations suggest that this 
is true, and that D. ordii may be utilizing areas with more cover. 

The greatest dietary similarity between D. merriami and D. ordii oc- 
curred on the quartz sand hills. Although the percent cover in this habitat is 
approximately the same (21%) as the surrounding creosote-bush flats, the 
sand hills support three times as many species as the flats (Burgess and 
Northington 1977). This indicates that the diets of these two species may be 
more similar in areas of greater diversity. The diets of these two kangaroo 
rats on the creosote-bush flats were more similar during August than during 
February. Although quantitative data on the availability of seeds aie not 
available, field notes of various workers in the area indicate that during 1974 
the plants set seed from late spring through the autumn. This further 



366 O'CONNELL 

suggests that greater abundance may lead to more similarity between the 
diets of these two species. The tendency of these two kangaroo rats to con- 
centrate on different proportions of the food resources during times of less 
abundance may be one mechanism allowing their coexistence. 

Both D. merriami and D. ordii appear to be opportunistic feeders, able to 
shift their diets in response to varying habitat conditions. The lowest 
similarity indexes obtained from all inter- and intraspecific comparisons 
were those comparing the same species from different habitats. Intraspecific 
comparisons of the diets of D. merriami and D. ordii between the quartz 
sand hills (area no. 5; Fig. 1) and the creosote-bush flats (areas no. 3, 4; Fig. 
1) yielded similarity indexes of 12 and 13, respectively. Thus, the diets of 
these two species may differ between regions and between years in the same 
area. Reichman (1975) has reported that D. merriami ate more seeds (78.4%) 
and less insects (15.5%) in a Ldrrrea-dominated area of southern Arizona 
than has been reported here. However, there are most likely certain 
heteromyid food-habit trends that are common among all North American 
deserts. 

Variation in Number of Food Items in Diets 

Emlen (1966, 1968) has proposed that animals will be more selective and 
specialized in their diets when food is abundant and will become less selective 
and specialized as food items become scarce. Smigel and Rosenzweig(1974) 
have demonstrated, with radioactively labeled commercial seeds, that D. 
merriami and Perognathus penicillatus are more selective at higher seed 
densities. 

The data presented here also support the model proposed by Emlen (1966, 
1968). For both D. merriami and D. ordii the number of kinds of food items 
ingested (food niche breadth) was significantly less during the summer and 
autumn than during the winter and early spring. The food-habit data show 
that at certain times of the year one food item, for example, Euphorbia 
during November, will be utilized almost exclusively, whereas throughout 
the rest of the year, this item enters into the diet with much less frequency. 

The comparison of the food items ingested by D. merriami among 
different months within one habitat reveals that a wide variety of food items 
were used during February. Although only one new food item appeared in 
the diet during August, seven items were not part of the diet. A similar situa- 
tion occurred in November. Two food items appearing in the November diet 
had not been part of the February diet, whereas eight items appeared in 
February but not in November. These data further indicate that D. merriami 
is less discriminating in February and more selective in August and 
November. 

Comparison of Cheek-Pouch and Stomach Contents 

Many workers who have examined food habits of different species of 
Dipodomys have based their conclusions solely on the analysis of 



COEXISTENCE OF DIPODOMYS 367 

cheek-pouch contents (Reynolds 1958; Dunham 1968; Chapman 1972; 
Gaby 1972). Gaby (1972) compared the diets, as determined by analysis of 
cheek-pouch contents, of D. merriami and D. ordii in southern New 
Mexico. Although the species of seeds eaten were generally similar to those 
found in this study, the proportions were different. Gaby reported neither 
green vegetation nor insects as important constituents of these kangaroo 
rats' diets. 

Although cheek -pouch contents do give an indication of what the kanga- 
roo rats are harvesting, the data presented in this paper demonstrate that 
cheek-pouch contents are unreliable indicators of relative proportions of 
foods in the diet. Percent composition of food items obtained from analysis 
of cheek-pouch contents varied greatly from that obtained from stomach 
analyses. The mean number of food items per cheek-pouch was lower and 
more uniform throughout the year than that of food items per stomach. The 
size of seeds in the cheek pouches were generally larger than those from the 
stomachs. Grass seeds were an exception, but they were usually found still on 
the spikelets. 

Reichman (1975) has suggested that the items collected in heteromyid 
cheek pouches are for potential use rather than immediate ingestion. The 
findings of this study support this view. Larrea seeds were more than twice as 
important in the cheek-pouch contents than in the stomach contents. These 
seeds were present in the cheek pouches in equal dry weights throughout the 
year, but the frequency of Larrea seeds in the stomach contents fluctuated. 
Larrea is the dominant plant in the study area, making up almost all of the 
21% total cover on the creosote-bush flats (Burgess and Northington 1977). 
The data suggest that the kangaroo rats are collecting the more abundant, 
more stable food sources for storage purposes, while immediately ingesting 
the less stable sources. 

The high percentage of seeds and insignificance of insects and green 
vegetation in the cheek pouches (Fig. 7) further support the suggestion that 
the contents of cheek pouches are intended for storage purposes. Seeds are 
much better suited for storage in surface caches or burrows than are either 
insects or green vegetation. The higher percentage of larger seeds and lower 
percentage of smaller seeds in the cheek pouches as compared to stomach 
contents indicate that ease in handling different seed sizes may also be a 
factor determining what is placed in the cheek pouches or eaten immediately. 

Relation of Dipodomys to the Guadalupe Mountains National Park 

The western portion of the Guadalupe Mountains National Park was 
heavily grazed from the early 1900s up until the formation of the National 
Park in 1972. Warnock (undated) and Burgess and Northington (1977) point 
to grazing as a factor causing the increase of Chihuahuan Desert vegetation 
over grassland species. The question which immediately arises is, now will 
the vegetation respond to the termination of grazing in the area, and how will 
this affect the local animal populations? Reynolds (1950) studied the ecology 



368 O'CONNELL 

of D. merriami on the grazing lands of southern Arizona. By comparing the 
population densities of this kangaroo rat on grazed and ungrazed plots, he 
concluded that D. merriami favors grazed areas and avoids areas of dense 
perennial grasses. Rosenzweig (1973) also reported that D. merriami prefers 
open habitats. Thus the grazing practices in the western portion of the 
Guadalupe Mountains National Park may have favored D. merriami. If 
grasses were to become reestablished on the west side of the park, as 
Warnock (undated) suggests, this might cause a decrease in the D. merriami 
population. The high percentage of grass seeds in the diet of D. ordii suggests 
that an increase in grass cover may favor this species. Further monitoring of 
Dipodomys in this area will yield valuable information on the response of 
these animals to changing range conditions. 

The activities of kangaroo rats also may have an effect on range condi- 
tions. Their habits of storing seeds in surface caches or burrows may effect 
the dispersal of plant species. Reynolds (1958) found that D. merriami has 
little effect on range conditions when the range is in good to excellent 
condition. Shrub seeds may be planted by the rodents, but under good range 
conditions their dispersal is slow. As range conditions deteriorate, he found 
numbers of D. merriami to increase and their presence to have an effect on 
range conditions. In advanced stages of range deterioration, Reynolds 
(1958) suggested that the activities of D. merriami may be sufficient to pre- 
vent range recovery even though cattle grazing is terminated. 

The range conditions of the western portion of the Guadalupe Mountains 
National Park probably fall in this last category. Thus the activities of D. 
merriami in the area may have a negative effect on range recovery. The 
analysis of cheek-pouch contents has shown that D. merriami is probably 
storing the larger seeds of shrubs and succulent plants that are often indica- 
tors of poor range conditions. Under deteriorated range conditions, the 
storing activities of D. merriami have been found to aid in the dispersal of 
these plants (Reynolds 1958). The establishment of fenced-in plots to study 
the response of range vegetation (1) to no kangaroo rats; (2) to the presence 
of one species; and (3) to the presence of both species would yield critical 
information on the effect of these rodents on range conditions. 

The 1974 Master Plan for the Guadalupe Mountains National Park (U.S. 
Department of Interior 1974) indicates that in the future the western portion 
of the park may be transversed by roads. The effects of roads on popula- 
tions of small mammals is not well known. Oxley et al. (1974) found that 
roads through northern coniferous forests act as barriers to the dispersal of 
small mammals. The open spaces of the roads and their edges inhibited the 
movements of the animals. In the more open desert regions, this effect 
caused by roads probably is not an important factor. Roads may increase the 
food supply as a result of increased moisture from run-off of rain water. The 
increased supply of grasses along roadsides has caused an expansion of the 
ranges of some microtine rodents (Baker 1971). A paved road through the 
western portion of the Guadalupe Mountains National Park may alter the 



COEXISTENCE OF DIPODOMYS 369 

vegetation adjacent to the road. An alteration of vegetation conditions may 
create changes in the distribution of rodent populations. For example, a 
road in this area may provide a corridor along which D. ordii could expand 
its distribution. If a road is built through this area, continued study of the 
rodent populations would provide a better understanding of the effects of 
roads on small mammals. 

The average visitor to the Guadalupe Mountains National Park seldom 
sees the park's many mammals because these animals are elusive and pri- 
marily nocturnal. However, the careful observer can find an abundance of 
mammal signs in the area. One of the purposes of an interpretive program in 
the National Parks should be to help the visitor become a better and more 
appreciative observer. Kangaroo rats are well suited for use in interpretive 
programs. Although seldom seen, these animals construct conspicuous 
burrows under mesquite and creosote bushes. Their distinctive tracks are 
often seen in the sandy washes and by the surface caches they dig for seed 
storage. In addition, kangaroo rats are relatively easy to keep in captivity for 
display purposes. A combination of live kangaroo rats, with an explanation 
of their adaptations to desert habitats, and an exhibit explaining these 
rodents' signs could be an instructive part of the park's interpretive program. 

The similarity indexes suggest that the diets of these two kangaroo rats are 
more similar with greater diversity and abundance of food resources. When 
food items become less diverse and abundant, the diets of the two species 
become less similar. The data suggest that in times of food scarcity the ani- 
mals will concentrate on different food resources or on different propor- 
tions of the same resources. If future activities of man in the area were to 
create conditions of stress for these kangaroo rats, this stress would be 
reflected in their diets. Thus the diets of Dipodomys merriami and D. ordii 
could be used not only as biological indicators of the human impact on 
diversity and abundance of the habitat but also of human impact on these 
rodents populations. 

LITERATURE CITED 

Baker, R. H. 1971. Nutritional strategies of myomorph rodents in North 

American grasslands. /. Mammal. 52:800-805. 
Baumgartner, L. L., and A. C. Martin. 1939. Plant histology as an aid in 

squirrel food-habit studies. /. Wildl. Manage. 3:266-268. 
Brown, J. H. 1973. Species diversity of seed-eating desert rodents in sand dune 

habitats. Ecology. 54:775-787. 
Brown, J. H., and G. A. Lieberman. 1973. Resource utilization and coexistence 

of seed-eating rodents in sand dune habitats. Ecology 54:788-797. 
Burgess, T. L., and D. K. Northington. 1977. Desert vegetation in the 

Guadalupe Mountains Region. In R. H. Wauer and D. H. Riskind, eds. 

Transactions — Symposium on the Biological Resources of the Chihuahuan 

Desert Region. National Park Service, Washington, D.C., in press. 
Chapman, B. R. 1972. Food habits of Loring's kangaroo rat, Dipodomys elator. J. 

Mammal. 53:877-880. 



370 O'CONNELL 

Chew, R. M., and B. B. Butterworth. 1964. Ecology of rodents in Indian Cove 

(Mojave Desert), Joshua Tree National Monument, California. J. Mammal. 

45:203-225. 
Dunham, M. K. 1968. A comparative food habit study of two species of kangaroo 

rate — Dipodomys ordii and D. merriami. M. S. Thesis. Univ. New Mexico, 

Albuquerque, 25 pp. 
Dusi, J. L. 1949. Methods for the determination of food habits by plant microtech- 
niques and histology and their application to cottontail rabbit food habits. /. 

Wildl. Manage. 13:295-298. 
Emelen, J. M. 1966. The role of time and energy in food preference. Am. Nat. 

100:611-617. 

1968. Optimal choice in animals. Am. Nat. 102:385-389. 

Flinders, J. T., and R. M. Hansen. 1972. Diets and habitats of jackrabbits in 

northeastern Colorado. Colorado State Univ. Range Sci. Dept. Sci. Ser., 

12:1-29. 
Fracker, S. B., and J. A. Brischle. 1944. Measuring the local distribution of 

Ribes. Ecology 25:283-303. 
Gaby, R. S. 1972. Comparative niche utilization by two species of kangaroo rats 

(genus Dipodomys). Ph.D. Thesis, New Mexico State Univ., Las Cruces, 71 pp. 
Genoways, H. H., R. J. Baker, and J. E. Cornely. 1977. Mammals of the 

Guadalupe Mountains National Park, Texas. This volume. 
Hansen, R. M., and J. T. Flinders. 1969. Food habits of North American hares. 

Colorado State Univ. Range Sci. Dept. Sci. Ser., 1:1-18. 
Hansson, L. 1970. Methods of morphological diet microanalysis in rodents. Oikos 

21:255-267. 
Hawbecker, A. C. 1951. Small mammal relationships in an ephedra community. 

J. Mammal. 32:50-60. 
MacArthur, R., and R. Levins. 1964. Competition, habitat selection and 

character displacement in a patchy environment. Proc. Natl. Acad. Sci. 

51:1207-1210. 
MacArthur, R., and E. R. Pianka. 1966. On optimal use of a patchy environ- 
ment. Am. Nat. 100:603-610. 
OOSTING, H. J. 1956. The Study of Plant Communities; An Introduction to Plant 

Ecology. W. H. Freeman and Co., San Francisco, 440 pp. 
OXLEY, D. J., M. B. Fenton, and G. R. Carmody. 1974. The effects of roads on 

populations of small mammals. J. Appl. Ecol. 11:51-59. 
Reichman, O. J. 1975. Relation of desert rodent diets to available resources. J. 

Mammal. 56:731-751. 
Reynolds, H. G. 1950. Relation of Merriam kangaroo rats to range vegetation in 

southern Arizona. Ecology 31:456-463. 
1958. Ecology of the Merriam kangaroo rat {Dipodomys merriami 

Mearns) on the grazing lands of southern Arizona. Ecol. Monogr. 28:1 1 1-127. 
Rosenzweig, M. L. 1973. Habitat selection experiments with a pair of coexisting 

heteromyid rodent species. Ecology 54: 1 1 1 - 1 1 7. 
Smigel, B. W., and M. L. Rosenzweig. 1974. Seed selection in Dipodomys 

merriami and Perognathus penicillatus. Ecology 55:329-339. 
Sokal. R. R., and F. J. Rohlf. 1969. Biometry. W. H. Freeman and Company, 

San F-ancisco, 776 pp. 



COEXISTENCE OF DIPODOMYS 371 

Sparks, D. R. and J. C. Malechek. 1968. Estimating percentage dry weight in 
diets using a microscope technique. J. Range Manage. 21:264-265. 

Ueckert, D. N., and R. M. Hansen, 1971. Dietary overlap in grasshoppers on 
sandhill rangeland in northeastern Colorado. Oecologia (Berl.), 8:276-295. 

United States Department of Commerce Publications. 1973, 1974, 1975. 
Climatological Data, Texas, vols. 78, 79, 80. 

United States Department of the Interior. 1974. Draft environmental state- 
ment and master plan, Guadalupe Mountains National Park. 

Warnock, B. H. undated. Plant communities of the Guadalupe Mountains in 
Texas and nearby Carlsbad Caverns National Park, unpubl. 



ACKNOWLEDGMENTS 

I gratefully acknowledge Dr. Robert J. Baker, Dr. Hugh H. Genoways, 
and Dr. Darrell N. Ueckert for their support in this project and their critical 
readings of the manuscript. Dr. Jerran T. Flinders assisted in the planning of 
the study. 

James W. Cottrell, John E. Comely, L. T. Green, Lloyd Logan, Jeanne E. 
Stone, Brent L. Davis, D. Craig Rudolf, Robert J. Baker, Janie B. Murray, 
and Walter W. Walthall aided in the collection of specimens. I thank L. T. 
Green and Molly A. Walker for help in preparation of specimens. Tony L. 
Burgess assisted in the identification of plants and Janie B. Murray helped in 
the preparation of the reference collection. 

I would especially like to thank Gary Ahlstrand, John Chapman, and 
Roger Reisch of the U.S. National Park Service for their cooperation and 
logistic support. 

This project was supported in part by U.S. National Park Service 
Contract CX700040145. 



Ecological Distribution of Woodrats 
(Genus Neotomaj in Guadalupe 
Mountains National Park, Texas 



JOHN E. CORNELY, Texas Tech University, Lubbock 

The preservation of our national parks for the enjoyment and enlighten- 
ment of future generations depends on maintenance of human impact at 
levels which will not degrade natural assemblages of plants and animals. The 
slogan of the National Park Service, "parks are for the people," challenges us 
to develop management plans that will permit maximum use of the parks 
with minimal impact on the local ecosystems. A thorough understanding of 
interrelationships and ecological requirements of a park's flora and fauna is 
a prerequisite to solving this difficult problem. Once these relationships are 
known, methods of measuring human impact on the ecosystem must be 
developed and employed to monitor the condition of the ecosystem. 

Although small mammals are an important part of the ecosystem, they 
have generally been overlooked in the development of resource manage- 
ment plans. Small mammal studies in our national parks have tended to be 
descriptive inventories without management implications. Some of the small 
mammalian species occurring in our national parks are endangered and 
should be actively managed to enhance their chances of survival. Small 
mammals also deserve careful study because of their role as prey for car- 
nivores, because they may compete with other animals for food resources, 
and because of their impact, both positive and negative, on the flora of a 
park. A final consideration is utilization of small mammals as biological 
indicators of habitat conditions within a park. 

Biological indicators are organisms, species, or communities which indi- 
cate the presence of certain environmental conditions. Plants have been 
widely used as indicators of habitat conditions and have proved useful as 
habitat management aids. Animals associated with these plants also can be 
utilized as habitat indicators. Human impact in a national park can be deter- 
mined in part by monitoring changes in quantity and quality of different 
types of habitats and accompanying changes in density and distribution of 
associated animals. 

373 



374 CORNELY 

The purpose of this research was to determine the distribution of wood- 
rats in Guadalupe Mountains National Park, to quantify the size and com- 
ponents of representative woodrat houses, to analyze the habitat in the 
immediate vicinity of these houses, and to investigate the feasibility of using 
woodrats as biological indicators of habitat conditions. 

Three species of woodrats occur in Guadalupe Mountains National Park. 
In 1901, Bailey (1905) collected Neotoma albigula in Dog Canyon. He also 
reported that N. mexicana was common at higher elevations in the 
Guadalupe Mountains of Texas, living in rocks and cliffs and ranging to the 
tops of the mountains. Davis (1940) trapped one N. albigula at Frijole in 
1938 and one N. mexicana in a log cabin in The Bowl in 1939. Davis and 
Robertson (1944) reported N. micropus from Culberson County, Texas, but 
did not indicate their presence in the immediate vicinity of the Guadalupe 
Mountains. All three species of woodrats were captured within the 
boundaries of Guadalupe Mountains National Park during my initial mam- 
mal survey work in 1973. 

METHODS AND MATERIALS 

The distribution of woodrats in Guadalupe Mountains National Park was 
mapped as the result of extensive trapping throughout the park in conjunc- 
tion with a survey of mammals. Woodrats were captured in Sherman folding 
aluminum live traps, wire-mesh traps ( 18 x 4>/2 x 5 in.), and Victor rat traps. 
Several woodrats were shot with a 0.22 caliber pistol. Specimens were 
deposited in The Museum, Texas Tech University. Most specimens were 
prepared as museum skins and skulls, but some were skulls without skins, 
complete skeletons without skins, and skulls accompanied by bodies pres- 
erved in alcohol. For descriptions and locations of collecting sites and ac- 
counts of species, see Genoways et al. (1977). Use of the terms "house," 
"den," and "nest" follows Finley (1958). A "den" is any large outer shelter 
enclosing the living area of the occupant; a "house" is a den constructed by 
the occupant; and a "nest" is a small resting place lined with soft fibrous 
material. 

Ten houses of Neotoma albigula and 10 of N. micropus were examined. 
Individuals of N. mexicana do not construct houses and were, therefore, 
excluded from this phase of the study. Only houses from which the resident 
woodrat had been collected were selected for analysis. House length was 
measured as greatest length, width at the widest point perpendicular to the 
length, and height was measured at the highest point. Each house was dis- 
mantled. Materials used to construct the house were separated and weighed 
in a canvas sling suspended from a Chatillon spring-balance (calibrated from 
to 1 5 kg). All materials of insufficient combined weight to be measured and 
all items apparently cached in a house were listed. Length, width, and height 
of the houses and weight of construction materials were analyzed by a one- 
way analysis of variance to test for significant differences among or between 
means (Sokal and Rohlf 1969). When means were significantly different, the 



WOODRAT DISTRIBUTION 



375 



Student-Newman-Keuls (SNK) procedure was used to determine maximal 
nonsignificant subsets (Sokal and Rohlf 1969). 

The vegetation around each house and around 10 den sites of Neotoma 
mexicana was sampled by means of a line-intercept method, modified from 
the procedure described by Canfield (1941). Using the nest chamber as the 
midpoint, two 10-m line-intercepts were established. One intercept was 
oriented north and south by compass and the other was oriented east and 
west (Fig. 1 ). A list of all plant species which occurred within the plot formed 
by connecting the ends of line-intercepts was recorded. From these lists, 
floral similarity indices were computed for paired comparisons between 
plots. The indices were computed using the formula for Pirlot's index 
(Mosimann 1968). 



/|\ 



/ nest 



E 

o 



\l/ 



Fig. 1. Design for analysis of vegetation around woodrat house 
or den. 



RESULTS 



Woodrat Distribution 

Neotoma mexicana probably occurs throughout the Guadalupe Moun- 
tains at elevations above 1500 m. Mexican woodrats have been collected at 
the southeastern base of the mountains in Bell Canyon (Davis and Robert- 



376 



CORNELY 




- % 



^ 












#-i 



y< 






Fig. 2. Distribution of woodrats in Guadalupe Mountains National Park, Texas. 



son 1944), on the walls and around the perimeter of Upper Dog Canyon, and 
in The Bowl (Fig. 2). 

Neotoma albigula (Fig. 3) occurs around the perimeter of the mountains 
and on the floors of Upper Dog and West Dog Canyons which penetrate the 
mountain mass. On the west side of the park, the white-throated woodrat is 
found primarily in or along edges of dry washes extending westward from 
the mountains. 

Neotoma micropus has been found only in the southern portion of the 
park. Specimens have been collected near Williams Ranch Road entrance on 
the southeastern boundary of the park, in a large dry wash which skirts the 
north edge of the central ridge of the Patterson Hills, and near Lewis Well in 
the extreme western area of the park. 

Neotcma mexicana and N. albigula are in contact around the perimeter of 
Upper Dog Canyon. In 1901 Bailey (1905) collected one specimen of N. 



WOODRAT DISTRIBUTION 



377 




Fig. 3. Adult female Neotoma albigula. 



albigula in Upper Dog Canyon at 2073 m (6800 ft). He also collected four 
specimens of N. mexicana, three at 2134 m and one at 2377 m. In 1973 both 
species were captured in a trap line that extended from the canyon floor up 
the eastern wall. Specimens of N. albigula were captured only on the canyon 
floor and specimens of N. mexicana were found from the base of the canyon 
wall on up the slope (Fig. 4). N. mexicana seemed to be the more abundant 
species. In 1974 no individuals of N. mexicana were collected below an eleva- 
tion of 2000 m, approximately 100 m above the canyon floor. Several speci- 
mens of N. albigula were collected 30 to 40 m above the canyon floor in 1975, 
indicating a possible upward shift in the zone of contact between the two 
species. Gehlbach (pers. comm.) and Scudday (pers. comm.) reported both 
N. mexicana and N. albigula in close proximity near Pratt Lodge in McKit- 
trick Canyon in the late 1950s. No woodrats were found during extensive 
trapping in McKittrick Canyon in 1973 and 1974. 

Neotoma micropus and N. albigula are in contact immediately north of 
the Patterson Hills (Fig. 5). North of this area only N. albigula has been 
captured and south of the area only N. micropus has been collected (Fig. 2). 
In a wash north of the road which extends due west of Williams Ranch 
House, several N. albigula were collected, but no N. micropus were found. In 
another wash 400 m to the south, one N. albigula was collected from a house 
that was central to several houses from which specimens of N. micropus were 
taken. 



378 CORNELY 



- 



v£?-V- ■ 




€2> 



.:*«. *s „• * j*.-,; . 




Fig. 4. (/ibove) Rock outcrop on the east wall of Upper Dog Canyon inhabited by 
Neotoma mexicana. 



Fig. 5. {Below) Area of contact between Neotoma albigula and Neotoma micropus 
immediately north of the Patterson Hills. 



WOODRAT DISTRIBUTION 379 

House Sites 

Houses inhabited by Neotoma albigula were found in a wide variety of 
sites. In Upper Dog Canyon a large number of houses had been constructed 
in association with fallen juniper trees or under living junipers. Houses were 
observed under red barberry {Berberis haematocarpa) and clumps of prickly 
pear {Opuntia lindheimeri) and cholla (Opuntia imbricata). Around the 
perimeter of Upper Dog Canyon and for a short distance up the slope, N. 
albigula occurs in rocks, utilizing rock crevices for shelter and filling the 
crevices with sticks, cactus, and debris rather than constructing a house. 
These rock dens resemble den sites of N. mexicana, but appear to have much 
more material, expecially cactus, stuffed into the crevices. In general, houses 
inhabited by N. albigula in Upper Dog occurred throughout the floor of the 
canyon wherever suitable shelter was available. 

In West Dog Canyon N. albigula houses appeared to be concentrated in or 
near the dry wash on the canyon floor. Houses were observed under large 
shrubs, such as red barberry, along the wash. One house was on the edge of 
the east bank of the wash with a subterranean entrance in the side of the bank 
about 0.5 m below the main house. All houses observed in West Dog Canyon 
were located in very dense, shrubby vegetation. 

In the northwest corner of the park, all N. albigula houses investigated 
were situated in large clumps of prickly pear. The cactus clumps were very 
dense and woodrats had piled cactus joints around the base of the clumps to 
form houses. This general form of house seems typical for N. albigula on the 
west side of the park. Large clumps of cactus most often occur in or along 
edges of dry washes on the we r t side, on slopes of the Patterson Hills, and 
isolated hills such as the Stage Coach Hills. Large clumps of Opuntia are not 
common on the creosote-bush bajada which dominates the west side of the 
park. N. albigula houses in large clumps of prickly pear were observed 
around the Williams Ranch House at the west base of the Guadalupe escarp- 
ment, at a locality northwest of the Ranch House and in the northwest 
corner of the park. 

Although Neotoma albigula houses are predominantly found under large 
growths of Opuntia on the west side of the park, where N. albigula was found 
in close proximity to N. micropus, the N. albigula house sites were quite 
different. In this area only one N. albigula was collected from a house con- 
structed under a large prickly pear. This house was central to several houses 
from which specimens of N. micropus had been collected. All of these houses 
were located in the large wash which skirts the north end of the central ridge 
of the Patterson Hills. In another wash approximately 400 m to the north, 
several N. albigula houses were investigated. Very few cacti are found in this 
wash and those that do occur are small and scattered. Every active house 
investigated in this wash was inhabited by an individual N. albigula, and all 
but one were situated under clumps of mesquite (Prosopis glandulosd\. The 
remaining house was in a dense growth of Brickellia laciniata, a shrubby 
composite. 



380 CORNELY 

On the east side of the park near Nipple Hill, N. albigula houses are 
located predominantly under prickly pear or cholla. 

A majority of the houses investigated that were inhabited by Neotoma 
micropus had been constructed under large clumps of prickly pear or cholla. 
Most sites were on the floor or along edges of dry washes. One N. micropus 
was collected from a house on a mesquite hummock near Lewis Well, and 
several apparently inactive houses were observed on other mesquite 
hummocks in this vicinity. 

Neotoma mexicana inhabits cliffs and rocks and does not construct a 
house. These woodrats construct nests in rock crevices and may deposit 
some plant debris and other materials in crevices, but not to the extent that is 
characteristic of individuals of N. albigula. In Upper Dog Canyon N. 
mexicana are found on rocky canyon walls. These woodrats seem to prefer 
vertical crevices for nest sites, but will utilize deep horizontal crevices with 
narrow openings. 

In The Bowl several Mexican woodrats have been collected in an old log 
cabin where a nest of shredded paper had been constructed in the far corner 
of an old bunk. A large number of acorns had been deposited around the nest 
and an old section of eaves spout was approximately one-third full of acorns. 
Specimens of N. mexicana were collected on a rock outcrop on a slope above 
an earthen dam in The Bowl. Woodrat signs were observed along rock out- 
crops near Bush Mountain and above the Blue Ridge campsite. Although no 
specimens were taken, these areas are probably inhabited by N. mexicana. 

House Analyses 

Because Neotoma albigula occurs in a variety of habitats within the park, 
houses from two distinctly different areas were selected for investigation. 
Five houses of N albigula were examined in Upper Dog Canyon and five 
were examined near the Crossroads immediately north of the Patterson 
Hills. Four of the houses examined in Upper Dog Canyon had been con- 
structed in association with large fallen juniper trees in an open woodland 
(Fig. 6), whereas the fifth house was located under a large, living alligator 
juniper (Juniperus deppeana) and was protected by large red barberry. 

Of the N. albigula houses investigated near the Crossroads, four were 
located in a dry wash immediately north of the Crossroads and a fifth was in 
the large wash which skirts the north edge of the central ridge of the Patter- 
son Hills. Three of these houses had been constructed under spreading, 
many-stemmed mesquite bushes (Fig. 7), one was in a dense growth of 
Brickellia laciniata, and one was under a large prickly pear. 

Ten houses of Neotoma micropus were investigated in the large wash 
which skirts the north edge of the central ridge of the Patterson Hills. Seven 
of these houses had been constructed under large prickly pears (Opuntia 
lindheimeri and O. phaecantha) (Fig. 8), one was under a cholla, and one was 
under a creosote bush (Larrea tridentata). 



WOODRAT DISTRIBUTION 



381 



4 

,4' 







« 




r *3fc» 






' !»^ ! 





















\*v*- - J*2»5!^*«j»t5£fafi&* 



*ifvW <; 



%. * : 



Fig. 6. (Above) Neotoma albigula house in association with a fallen juniper in Upper 
Dog Canyon. 



Fig. 7. (Below) Neotoma albigula house under mesquite in a dry wash immediately 
north of the Crossroads. 



382 CORNELY 




Fig. 8. Neotoma micropus house under large prickly pear in a large wash which 
skirts the north edge of the central ridge of the Patterson Hills. 



The results of the quantification of house size and weights of materials 
used in constructing the house are summarized in Table 1. The means are 
relatively similar except for the mean weights of sticks used in den con- 
struction. Analyses of variance (ANOVA) of house length, width, and 
height, and weights of cholla used in construction revealed no significant 
differences between or among species or localities. Results of ANOVA of 
weights of sticks in houses are summarized in Table 2. The ANOVA yielded 



TABLE 1 . Mean house measurements and mean weights of construction components of wood- 
rat houses in Guadalupe Mountains National Park. 





N. albigula (n = 5) 


N. albigula (n - 5) 


N. micropus (n - 10) 




Upper Dog Canyon 
Mean SE 


Crossroads 


Crossroads 


Measurements 


Mean SE 


Mean SE 


House length 


158.4 cm (19.8115) 


106.8 cm (12.1507) 


133.5 cm (13.6943) 


House width 


75.2 cm (12.9711) 


82.2 cm (8.1704) 


84.2 cm (6.4735) 


House height 


37.7 cm (6.5019) 


25.9 cm (1.7493) 


29.15 cm (3.1056) 


Sticks 


9.66 kg (3.1103) 


3.45 kg (0.5508) 


2.1 kg (0.5888) 


Cholla 


0.99 kg (0.3509) 


0.26 kg (0.2358) 


1.96 kg (0.6704) 


PrickW pear 


0.00 kg (0.00) 


0.02 kg (0.0120) 


0.05 kg (0.0354) 


Manure 


0.45 kg (0.1175) 


0.05 kg (0.0316) 


0.21 kg (0.1234) 



WOODRAT DISTRIBUTION 



383 



TABLE 2. Results of analysis of variance (ANOVA) and Student-Newman-Keuls test (SNK) 
on weights of sticks in houses of Neotoma micropus and Neotoma albigula. Means are in 
kilograms. 

ANOVA Table 



Source 


d.f. 




MS 


F 


Among 
Within 


2 
17 




97.7790 

13.7255 


7.1239** 


SNK Test 







Rank 






1 


2 


3 


Mean 


2.10 


3.46 


9.66 


n 


10 


5 


5 



Rank Mean 


n 






1 JV. micropus 2.10 

2 N. albigula-Crossroads 3.46 

3 N. albigula-Upper Dog 9.66 


10 

5 
5 


1.36ns 

7.56** 


6.20** 


ns = Nonsignificant; ** = PK. 0.01. 









TABLE 3. Materials found in woodrat houses and number of houses in which they were found. 



Materials 



TV. albigula 
Upper Dog 



N. albigula 
Crossroads 



N. micropus 
Crossroads 

(Al=10) 



Sticks 

Cholla 

Tasajillo 

Prickly pear pad 

Prickly pear fruit 

Manure 

Juniper leaves 

Juniper berries 

Acorns 

Barberry leaves 

Pine needles 

Pine cones 

Century plant leaves 

Mesquite leaves 

Mesquite pods 

Bones 

Ocotillo 

Yucca pod 

Feathers 

Old undershorts 

Rabbit's foot 



5 


5 


5 


2 





1 


2 


2 








5 


4 


3 





2 





4 





2 





1 





1 





2 


1 





1 


2 


4 


2 














1 





1 





1 





1 



384 



CORNELY 



a highly significant F value of 7.1239**. A Student-Newman-Keuls pro- 
cedure (Table 2) revealed that significantly greater weights of sticks were 
found in Neotoma albigula houses in Upper Dog Canyon than in either N. 
albigula or N. micropus houses near the Crossroads. As shown in Table 2, 
the homogeneous subsets seem to be correlated with locality rather than with 
species. 

The frequency of occurrence of materials accumulated in 20 dismantled 
houses is presented in Table 3. Note that sticks were found in every house 
and manure and cactus parts were quite common. Mesquite leaves and 
pods and tasajillo joints (Opuntia leptocaulis) were common in houses near 
the Crossroads, but absent from houses in Upper Dog Canyon. Juniper 
parts and acorns were commonly found in houses in Upper Dog Canyon, but 
were absent in houses from near the Crossroads. 

Vegetation Analyses 

Floral similarity indices were calculated for the paired comparison of 
vegetation in the immediate vicinity of each house or den to that of each 



TABLE 4. Mean floral resemblance indices between and within species and localities. 



Species compared 



Mean index 



SE 



N. albigula 
N. albigula 

N. albigula- 
N. albigula- 

N. albigula- 
N. albigula- 



-Upper Dog Canyon 
-Crossroads 



Upper Dog Canyon 
Upper Dog Canyon 



N. albigula — Crossroads 
N. albigula — Crossroads 

N. mexicana 
N. mexicana 

N. micropus 
N. micropus 

N. albigula— Upper Dog Canyon 
N. mexicana 

N. albigula — Upper Dog Canyon 
N. micropus 

N. albigula — Crossroads 
N. mexicana 

N. albigula — Crossroads 
N. micropus 

N. mexicana 
N. m ; cropus 



19.9244 


3.0848 


2.7760 


0.7900 


40.7840 


3.3505 


41.9360 


3.1781 


59.8078 


1.1697 


45.9142 


1.9564 


12.6594 


1.4474 


4.0648 


0.6239 


0.7670 


0.2856 


38.4232 


1.5246 


1.2034 


0.2415 



WOODRAT DISTRIBUTION 385 

other house or den. Mean indices between and among species are 
summarized in Table 4. Note the very low similarity between Neotoma 
albigula plots in Upper Dog and N. albigula plots near the Crossroads. This 
resulted in a mean index among all N. albigula plots that was much lower 
than mean indices among all N. mexicana plots and among all N. micropus 
plots. Because of low similarity between N. albigula plots in Upper Dog 
Canyon and N. albigula plots near the Crossroads, plots from these two 
localities were considered separately. The mean similarity index between N. 
micropus plots and N. albigula plots near the Crossroads is much higher 
than the mean index between N. albigula plots in Upper Dog Canyon and 
near the Crossroads. 

A modified line-intercept technique was developed to estimate coverages 
of plant species in the vicinity of a woodrat house, because the floral simi- 
larity index does not reflect frequency or coverage of a particular species 
within a plot, but merely its presence. To standardize the technique, the nest 
chamber was used as the midpoint of each transect. Traditional vegetation 
analysis methods are constructed to obtain a random sample; however, the 



TABLE 5. Dominant plants on woodrat house or den plots, estimated percent coverages of 
plant species on the plots, and percent of sampled plots on which the dominants were found. 



Species 


% coverage 


Frequency (%) 


Neotoma mexicana (n - 10) 






Nolina micrantha 


11.08 


80 


Muhlenbergia pauciflora 


9.595 


100 


Dasylirion leiophyllum 


7.985 


80 


Cercocarpus montanus 


7.19 


90 


Quercus undulata 


3.575 


80 


Neotoma albigula — Upper Dog Canyon (n - 5) 






Stipa tennuissima 


23.17 


80 


Bouteloua gracilis 


3.11 


80 


Xanthocephalum sarothrae 


2.86 


100 


Lycurus phleiodes 


2.5 


80 


Muhlenbergia repens 


1.64 


60 


Neotoma albigula — Crossroads (n - 5) 






Prosopis glandulosa 


17.32 


60 


Larrea tridentata 


8.38 


80 


Brickellia laciniata 


6.63 


60 


Muhlenbergia porteri 


6.48 


80 


Set aria leucopila 


0.79 


80 


Neotoma micropus (n = 10) 






Larrea tridentata 


17.39 


100 


Opuntia lindheimeri 


9.71 


50 


Prosopis glandulosa 


9.54 


50 


Muhlenbergia porteri 


5.94 


60 



386 CORNELY 

method developed for this study was standardized so that each plot would be 
sampled in the same manner. It should be noted here that the house is not 
necessarily the center of the resident woodrat's home range. Mean percent 
coverages of dominant plant species on plots are listed in Table 5. The 
percent of plots on which the species were found is also listed in Table 5. A 
plant was considered to be dominant only if it was present in 50% or more of 
the plots. Note that plots around N. micropus houses and those of N. albigula 
near the Crossroads have similar dominants with somewhat different 
coverages. The presence of Opuntia lindheimeri as a dominant on N. 
micropus plots in contrast to the absence of that plant on N. albigula plots is 
significant. 

In addition to ground coverage, understory cover and canopy cover were 
also measured. It is theoretically possible, therefore, to record coverages 
greater than 100%. These results must be interpreted as estimates of plant 
coverages in relationship to house or den sites and not as random samples of 
the vegetation of the area. Plots around houses of N. albigula in Upper Dog 
Canyon and near the Crossroads were considered separately because of their 
low floral similarity indices. 

DISCUSSION 
Woodrat Distribution 

The ranges of Neotoma mexicana, N albigula, and N. micropus overlap 
in Texas from the Guadalupe Mountains and Davis Mountains westward 
(Hall and Kelson 1959). However, geographic range descriptions of 
Neotoma can be misleading, because where two or more species of wood- 
rats occur in the same area, they often establish distinct zones of contact with 
little or no overlap. The distribution of woodrats in Guadalupe Mountains 
National Park conforms to this pattern (Fig. 2). Throughout the park the 
woodrat distribution is best described as microallopatry, with narrow zones 
and areas of microsympatry. Similar situations have been described in other 
areas where ranges of two or more species of woodrats overlap (Bailey 1905, 
1931; Finley 1958; Cameron 1971; Wright 1973). Reasons for this habitat 
partitioning are complex and vary depending upon species involved and the 
nature of the habitat. 

Finley (1958) reported that ecological distribution of woodrats is pri- 
marily determined by climbing ability, house construction ability, and diet. 
In addition, water economy probably influences distribution (Lee 1963; 
Boice and Boice 1968; Boice 1969; Birney and Twomey 1970). 

According to Finley (1958), Neotoma mexicana is the most agile climber 
of the woodrats species in Guadalupe Mountains National Park, followed in 
ability by N. albigula and N. micropus. Distributional patterns observed in 
the park correlate with these reported differences in climbing ability, with N. 
mexicana inhabiting steep cliffs and rocky slopes, N. albigula occurring 
from gentler rocky slopes out onto the flats, and N. micropus found only on 
the flats. 



WOODRAT DISTRIBUTION 387 

House-building activities are apparently correlated with the collecting 
instinct of woodrats. Finley (1958) reported that N. albigula and N. 
micropus exhibit strong collecting instincts and are capable of constructing 
large houses. The collecting instinct of N. mexicana is much weaker rela- 
tive to that of the other two species, and individuals of this species do not 
build houses. Where individuals of N. albigula lived in rock crevices in Up- 
per Dog Canyon, large amounts of sticks, cactus joints, and other materials 
have been carried to the den site and were stuffed into crevices. Rock dens 
of N. mexicana in the same area have very little accumulated material. 

Food-habit studies (Vorhies and Taylor 1940; Spencer and Spencer 1941; 
Finley 1958; Wood 1969) suggest that N. albigula and N. micropus utilize 
cacti and other succulents to a large extent for food. Finley (1958) reported 
that N. mexicana apparently dislikes cactus. 

Interrelationships between desert-dwelling species of woodrats and 
species of cactus (Opuntia spp.) have been well documented (Vorhies and 
Taylor 1940; Spencer and Spencer 1941; Vorhies 1945; Finley 1958; Lee 
1963; Raun 1966; Brown et al. 1972). Opuntia spp. are utilized for shelter, 
food, and water. Woodrats must depend on vegetation as a source of water 
because they cannot subsist on metabolic water (Schmidt-Nielsen et al. 1948; 
Schmidt-Nielsen and Schmidt-Nielsen 1952). Opuntia is an excellent water 
source because it has a high water content throughout the year (Lee 1963) 
and its cell sap has a low osmotic pressure (Korstian 1924). 

House Analyses 

Analyses of representative houses of Neotoma albigula and N. micropus 
support preliminary observations that house size and weights of materials 
used to construct a house are more dependent on availability of materials 
than on differences in degree of collecting instinct between the two species. 
Finley (1958) reported that houses of N. albigula and N. micropus found in 
similar habitats were indistinguishable. The results of this study substanti- 
ate Finley's observation. The significant difference in weights of sticks in N. 
albigula houses in two diverse habitats reflects the opportunistic nature of 
this species. Differences can probably be attributed to availability of juniper 
sticks in Upper Dog Canyon where most houses are located in an open juni- 
per woodland. 

Although no significant differences were revealed between houses of 
Neotoma albigula and N. micropus where the two species are micro- 
sympatric, there was a striking difference in house sites. Only one N. albigula 
house was located under a cactus in the zone of contact, whereas 80% of the 
N. micropus houses had been constructed under cactus (Opuntia spp.). 
Olsen (1973) postulated that shelter-site selection is based on the criterion of 
cover near the ground, and he found no significant response to stem types. 

It is notable that, although cholla was present on only two of the five N. 
albigula plots in Upper Dog Canyon, all houses contained cholla joints 
(Table 3). Although woodrats can drag cholla joints without much trouble, 



388 CORNELY 

they have difficulty handling large prickly pear pads. Prickly pear pads are 
found in greatest numbers in houses which have been constructed at the base 
of a living prickly pear. Pads found in other houses were always small in size. 
In areas where tasajillo was present, joints and fruits from this cactus were 
present in 90% of the houses. 

Vegetation Analyses 

The floral similarity indices computed for this study proved to be useful 
for preliminary analyses. A low index may indicate that two areas are so dis- 
similar that no additional analysis is required, but if the index is high, addi- 
tional statistical techniques may be employed to further analyze the data. An 
examination of the similarity indices suggested that it would be best to con- 
sider the N. albigula plots in two different localities separately (Table 4). 
Vegetation on house plots of N. albigula at the Crossroads closely resembled 
vegetation on house plots of N. micropus in the same area, but exhibited 
little resemblance to plots of N. albigula in Upper Dog Canyon. These results 
are consistent with results of house analyses. With the exception noted 
above, all indices between species were much lower than indices within 
species. Although floral similarity does not reflect the frequency or cover- 
age of plant species, it must be noted that the mere presence or absence of a 
plant species may be more important than its relative abundance. 

The modified line-intercept method designed for this study facilitates the 
description of vegetation in the immediate vicinity of a woodrat house or 
den. Standardization of the method allows comparison of vegetation sur- 
rounding different houses. Use of perpendicular line-intercepts tends to 
emphasize plants in the center of the plot and, therefore, emphasizes the 
plant selected for a house site. This bias was designed into the method 
because the plant chosen for a house site often provides shelter, food, and 
water and, therefore, may be the most important plant in the life of the resi- 
dent woodrat. A quadrat method of vegetation analysis was tested during 
this study, but was not considered to be nearly as satisfactory as the 
line- intercept method. 

Woodrat Ecology 

Finley (1958) suggested that Neotoma mexicana and N. albigula are suf- 
ficiently divergent ecologically to limit interspecific competition. Vegeta- 
tion data from this study indicate that these two species are found in quite 
different habitats in Guadalupe Mountains National Park in spite of the fact 
that they are in close contact around the perimeter of Upper Dog Canyon. 
The area of microsympatry generally coincides with an ecotone between the 
chaparral-like vegetation of the canyon wall and the open canyon woodland 
of the canyon floor. In the absence of N. mexicana, N. albigula could 
probably inhabit much of the area now occupied by N. mexicana, with the 
possible exception of vertical cliffs. N. mexicana would not invade the 
canyon floor to a great extent because of the limited number of rocky den 



WOODRAT DISTRIBUTION 389 

sites. Finley (1958) reported that an individual N. mexicana will oc- 
casionally invade an unoccupied N. albigula house near a rocky slope, but 
there is no evidence that Mexican woodrats ever construct a house. The 
partitioning of habitat in Upper Dog Canyon apparently is influenced 
strongly by differences in climbing ability, house-building ability, and 
vegetation requirement or preferences. 

There are contrasting reports on the preferred habitat of Neotoma 
albigula. Bailey (1905) never found N. albigula away from rocky situations 
in west Texas and described the species as a cliff dweller. Vorhies and Taylor 
(1940), on the other hand, found N. albigula in almost every habitat type in 
the Lower Sonoran zone in Arizona. They were common in the Upper 
Sonoran zone and present in the Transition zone up to 7000 ft with one 
specimen taken at 8200 ft in the Santa Rita Mountains. Bailey reported that 
N. albigula in west Texas apparently belonged to the Upper Sonoran zone, 
but extended into the Lower Sonoran along cliffs and rocky gulches. A pre- 
liminary examination of N. albigula distribution in Guadalupe Mountains 
National Park would lead to the same conclusion. The key to this paradox is 
probably the presence of N. micropus in west Texas, whereas the distribu- 
tion of this species does not extend into Arizona. 

Finley (1958) could not distinguish between ecological requirements of 
Neotoma albigula and N. micropus and concluded that they competed for 
house sites wherever they came in contact. He thought that the two species 
could coexist at low population levels with little competition, but that with 
higher populations, competition could become intense. Wright (1973) con- 
cluded that competitive exclusion of N. micropus by N. albigula may be 
occurring where the two species coexist in the Mesilla Valley of New Mexico. 

Neotoma albigula occurs in a variety of habitats in Guadalupe Moun- 
tains National Park and appears to have less specific habitat requirements 
than N. mexicana or N micropus. Where N. albigula and N. micropus are in 
contact on the west side of the park, N. albigula has apparently shifted to a 
secondary habitat and diet. Diet studies (Vorhies and Taylor 1940; Spencer 
and Spencer 1941; Finley 1958; Wood 1969) suggest that both N. albigula 
and N. micropus utilize cacti to a large extent for food. Field observations 
durting this study support those suggestions with one significant exception. 
Cacti were apparently being used for food at every N. micropus house 
investigated as evidenced by an abundance of partially eaten cactus joints 
and fruits. This same statement can be made for every N. albigula house 
observed except those at the zone of contact with N. micropus. Although the 
mean floral similarity index is relatively high between house plots of the two 
species in this area, the relative percent coverages are different. There is one 
important difference — Opuntia was present at only one of five N. albigula 
sites examined at the zone of contact. The one plot with Opuntia present was 
located in the midst of several N. micropus houses. There was no indication 
whether the house was constructed by the resident N. albigula or whether it 
had moved into a vacant N. micropus house. In contrast to this area of 



390 CORNELY 

microsympatry, all other specimens of N. albigula collected on the west side 
of the park were captured near houses constructed under growths of 
Opuntia. Where N. albigula and A r . micropus are in close proximity, N. 
albigula houses were located predominantly under mesquite. This apparent 
shift in habitat selection must be accompanied by a shift in diet. It appears 
that N. albigula utilizes cactus for food throughout the park except at the 
zone of contact with N micropus where cactus are almost totally absent 
from TV. albigula house sites. 

The observations discussed above support the principle of competitive 
exclusion postulated by Gause (1934). Neotoma albigula may be avoiding 
direct competition with the larger, and possibly more aggressive N. micropus 
by shifting to a secondary habitat and diet. A similar situation has been 
described with TV. lepida and N.fuscipes in California (Cameron 1971). It is 
quite possible that behavioral differences exist between N. albigula and N. 
mexicana that may be preventing N. albigula from invading areas where N. 
mexicana is present. 

In summary, the distribution of Neotoma albigula in Guadalupe Moun- 
tains National Park may be limited more by the presence of the other two 
species of woodrats than by habitat limitations. The distribution of N. 
mexicana is limited by availability of favorable habitat, and the distribution 
of N. micropus is limited by the presence of N. albigula and availability of 
favorable habitat. 

Ecological observations and data from this study are consistent with cur- 
rent systematic interpretations concerning these three species. Neotoma 
albigula and N micropus are considered to be closely allied (Anderson 1969; 
Finley 1958; Birney 1973). N. mexicana is placed in the same subgenus 
{Neotoma) as N. albigula and N. micropus (Goldman 1910), but is not as 
closely related to the other two species as they are to each other. N. micropus 
is considered intermediate between N.floridana and N. albigula (Anderson 
1969; Finley 1958; Birney 1973) and hybrids are known between N.floridana 
and N. micropus and between N micropus and N. albigula. Several wood- 
rats live-trapped in zones of contact during this study were karyotyped, but 
no evidence of hybridization was found. All three species found in Guada- 
lupe Mountains National Park can be distinguished karyotypically (Baker 
and Mascarello 1969). 

Woodrats comprise an important component of the ecosystem in Guada- 
lupe Mountains National Park. They are among the most abundant mam- 
malian species in the park and their role is complex. The following verte- 
brates that occur in the park have been reported to prey upon woodrats: 
snakes, hawks, owls, roadrunner, skunks, badger, gray fox, ringtail, coyote, 
and bobcat (Bailey 1931; Vorhies and Taylor 1940; Linsdale and Tevis 1951; 
Raun 1966). Vorhies and Taylor (1940) and Spencer and Spencer ( 1941) con- 
cluded that woodrats consume only small amounts of grass material. 
Woodiats often have been blamed for the spread of cactus, but there is no 
evidence to support this accusation. To the contrary, a cactus plant selected 



V/OODRAT DISTRIBUTION 391 

as a site for a woodrat house is apparently harmed by debris collected by the 
rat (Vorhies and Taylor 1940). Spencer and Spencer (1941) reported wood- 
rats tend to restrict the spread of cholla. My field observations do not sug- 
gest that cactus populations are being harmed to any great extent by wood- 
rat activity. Woodrats are an important food source for carnivores in 
Guadalupe Mountains National Park. They apparently have a minimal 
effect on the flora of the park and do not pose a serious threat to the food 
resources of other animals. 

Woodrats as Biological Indicators 

One of the primary objectives of this study was to determine the feasi- 
bility of using woodrats as biological indicators. Populations of woodrats 
reflect changes in the vegetation with which they are associated. Numbers of 
Neotoma albigula tend to increase as the result of overgrazing (Vorhies and 
Taylor 1940). Heavy grazing may promote growth of Opuntia and mesquite 
by reducing competition from grasses and by reducing frequency of fires by 
limiting the amount of fuel. Increases in Opuntia and mesquite would pro- 
vide more woodrat shelter sites. Because N. micropus prefers similar shelter 
sites, a similar response to overgrazing would be expected. The decrease in 
grazing pressure that accompanied the creation of Guadalupe Mountains 
National Park may result in a decrease in numbers of N. albigula and N. 
micropus. 

Raun (1966) reported that a large-scale die-off of prickly pear was accom- 
panied by disappearance of a Neotoma micropus population which used 
cactus for shelter. Loss of vigor and rotting of cactus apparently were caused 
by 4 years of abnormally high rainfall. Wright (1973) postulates that range 
extension by N albigula at the expense of N. micropus in the Mesilla Valley 
of New Mexico is in response to successional changes in vegetational com- 
position as a result of human activity. Vorhies and Taylor (1940) refer to TV. 
albigula as an "animal weed." 

It would be feasible to utilize woodrats as a biological indicator in Guada- 
lupe Mountains National Park. Shifts in the zones of contact between 
species could be detected easily and would indicate a change in habitat con- 
ditions favoring one species over the other. A continued warming trend over 
a number of years could result in the reduction of Neotoma mexicana 
habitat on the walls of Upper Dog Canyon. This would facilitate the inva- 
sion of N. albigula. The increase in the quality and quantity of grassland in 
the park should result in reduced abundance of cacti and, therefore, reduced 
numbers of N. micropus and N. albigula. The grassy meadows in Upper Dog 
Canyon are fragile. Increased human use in this area will probably damage 
the grassland and result in increased numbers of cacti and, consequently, 
increased numbers of TV. albigula. Range extensions and increased numbers 
of N. albigula may indicate general habitat degradation in the park This 
may be caused by human impact, climatic changes, or may be a result of 
wildlife activity. Species of cactus are severely harmed by fire (Dwyer and 



392 CORNELY 

Pieper 1967; Wright 1972; Heirman and Wright 1973), thus an increase in 
frequency of fire, whether natural or by prescription burning, would reduce 
the populations of N. micropus and N. albigula. 

Although it is evident that woodrats could be useful biological indicators, 
a combination of monitoring systems would be desirable to detect habitat 
changes and assess human impact. The monitoring of changes in the vege- 
tation should be of highest priority. Our knowledge of effects of vege- 
tational changes on mammalian populations is limited and this information 
is necessary in order to make meaningful resource management decisions. 
Once we know how changes in vegetation are reflected in mammal popula- 
tions, the distribution and abundance of mammalian species can be pre- 
dicted from vegetational analyses. The small mammal populations should be 
periodically monitored, and vegetational changes should be correlated with 
changes in small mammal demography. To implement such a program, a 
system of environmental impact monitoring could be designed to include a 
series of permanent grids for monitoring small mammal populations com- 
bined with a plan for monitoring the vegetation on each grid. Data resulting 
from periodic utilization of this system would be invaluable in assessing 
human impact, ramifications of changing climatic conditions, and responses 
to prescribed burning. 

LITERATURE CITED 

Anderson, S. 1969. Taxonomic status of the woodrat, Neotoma albigula, in 
southern Chihuahua, Mexico. Pages 25-50 in J. K. Jones, Jr., ed., Contributions 
in Mammalogy, Misc. Publ. Mus. Nat. Hist., Univ. Kans. 51:1-428. 

Bailey, V. 1905. Biological survey of Texas. N. Am. Fauna 25:1-222. 

1931. Mammals of New Mexico. N. Am. Fauna 53:1-412. 

Baker, R. J., and J. T. Mascarello. 1969. Karyotypic analyses of the genus 
Neotoma (Cricetidae, Rodentia). Cytogenetics 8:187-198. 

Birney, E. C. 1973. Systematics of three species of woodrats (genus Neotoma) in 
central North America. Misc. Publ. Mus. Nat. Hist. Univ. Kans. 58:1-173. 

Birney, E. C, and S. L. Twomey. 1970. Effects of sodium chloride on water con- 
sumption, weight, and survival in the woodrats, Neotoma micropus and 
Neotoma floridana. J. Mammal. 51:372-375. 

BoiCE, R. 1969. Water intake as a function of ease of access in Neotoma. J. Mam- 
mal. 50:605-607. 

BoiCE, R., and C. BoiCE. 1968. Water intake following capture and deprivation in 
southwestern rodents. Psychon. Sci. 12:104. 

Brown, J. H., G. A. Lieberman, and W. F. Dengler. 1972. Woodrats and 
cholla: dependence of a small mammal population on the density of cacti. 
Ecology 53:310-313. 

Cameron, G. N. 1971. Niche overlap and competition in woodrats. J. Mammal. 
52:288-296. 

Canfield, R. H. 1941. Application of the line-intercept method and sampling 
range vegetation. J. For. 39:388-394. 

Davis, W. B. 1940. Mammals of the Guadalupe Mountains of western Texas. Oc- 
cas. Pap. Mus. Zool. La. State Univ. 7:69-84. 



WOODRAT DISTRIBUTION 393 

Davis, W. B., and J. L. Robertson, Jr. 1944. The mammals of Culberson Coun- 
ty, Texas. /. Mammal. 25:254-273. 

Dwyer, D. D., and R. D. Pieper. 1967. Fire effects of blue grama-pinyon-juniper 
rangeland in New Mexico. J. Range Manage. 20:359-362. 

Finley, R. B., JR. 1958. The wood rats of Colorado: distribution and ecology. 
Univ. Kans. Publ. Mm. Nat. Hist. 10:213-552. 

GAUSE, G. F. 1934. The Struggle for Existence. Williams & Wilkins Co., 
Baltimore, 163 pp. 

Genoways, H. H., R. J. Baker, and J. E. Cornely. 1977. Mammals of Guada- 
lupe Mountains National Park, Texas. This volume. 

Goldman, E. A. 1910. Revision of the wood rats of the genus Neotoma. N. Am. 
Fauna 31:1-124. 

Hall, E. R., and K. R. Kelson. 1959. The mammals of North America. Ronald 
Press, New York. Vol. 2. 

Heirman, A. L., and H. A. Wright, 1973. Fire in medium fuels of West Texas. J. 
Range Manage. 26:331-335. 

Korstian, C. F. 1924. Density of cell sap in relation to environmental conditions 
in the Wasatch Mountains of Utah. J. Agric. Res. 28:845-907. 

Lee, A. K. 1963. The adaptations to arid environments in wood rats of the genus 
Neotoma. Univ. Calif. Publ. Zool. 64:57-96. 

Linsdale, J. M., and L. P. Tevis. 1951. The Dusky-Footed Wood Rat. Univ. 
Calif. Press, Berkeley, 664 pp. 

Mosimann, J. E. 1968. Elementary Probability for the Biological Sciences. 
Appleton-Century-Crofts, New York, 255 pp. 

Olsen, R. W. 1973. Shelter-site selection in the white-throated woodrat, Neotoma 
albigula. J. Mammal.S4:594-6\0. 

Raun, G. G. 1966. A population of wood rats (Neotoma micropus) in southern 
Texas. Bull. Tex. Mem. Mus. 11:1-62. 

Schmidt-Nielsen, K., and B. Schmidt-N ielsen. 1952. Water metabolism of desert 
mammals. Physiol. Rev. 32:135-166. 

Schmidt-Nielsen, B., K. Schmidt-Nielsen, A. BROKAW,and H. Schneider- 
man. 1948. Water conservation in desert rodents J. Cell Comp. Physiol. 
32:331-360. 

Sokal, R. R., and F. J. Rohlf. 1969. Biometry. W. H. Freeman Co., San Fran- 
cisco, 776 pp. 

Spencer, D. A. and A. L. Spencer 1941. Food habits of the white-throated 
wood rat in Arizona. /. Mammal. 22:280-284. 

Vorhies, C. T. 1945. Water requirements of desert animals in the southwest. Tech. 
Bull. Agric. Exp. Stn. Univ. Ariz. 107:487-525. 

Vorhies, C. T. and W. P. Taylor. 1940. Life history and ecology of the white- 
throated woodrat, Neotoma albigula albigula Hartley, in relation to grazing in 
Arizona. Tech. Bull. Agric. Exp. Stn. Univ. Ariz. 86:455-529. 

Wood, J. E. 1969. Rodent populations and their impact on desert rangelands. Bull. 
Agric. Exp. Stn. N. M. State Univ. 555:1-17. 

Wright, H. A. 1972. Shrub response to fire. Pages 204-217 in Wildland and 
Shrubs — Their Biology and Utilization. U.S. Dep. Agric. Forest Ser. Gen. Tech. 
Rep. INT-l,494p. 

Wright, M. E. 1973. Analysis of habitats of two woodrats in southern New Mex- 
ico. J. Mammal. 54:529-535. 



394 CORNELY 

ACKNOWLEDGMENTS 

I am grateful to Robert J. Baker, Hugh H. Genoways, David K. 
Northington, and Henry A. Wright for guidance and for critically reviewing 
this manuscript. Jerran T. Flinders made helpful suggestions in the plan- 
ning stages of the project. The field assistance of Dallas E. Wilhelm, 
Margaret A. O'Connell, James W. Cottrell, Tony L. Burgess, L. T. Green, 
Timothy Holland, and Brent L. Davis is gratefully acknowledged. I thank 
Gary Ahlstrand, John Chapman, Roger Reisch, and Phil Van Cleave for 
their outstanding logistic support and cooperation. Special thanks to my 
wife, Bea, who typed the manuscript and provided moral support. Financial 
support for this project was provided by National Park Service Contract 
CX700040145. 



Status of the Guadalupe Mountain 
Vole, Microtus mexicanus 
guadalupensis 



DALLAS E. WILHELM, JR., Texas Tech University, 
Lubbock 



Ecologists in recent years have been involved increasingly in programs 
attempting to define and assemble management programs for areas of 
human impact on environments that contain esthetically pleasing asso- 
ciations of plants, animals, and physiography. Preservation of ecosystems in 
their natural state generally becomes more difficult with increasing human 
activity. These efforts have been concerned largely with the biotic compo- 
nents that are most visible to the human visitor — the plants and the larger 
diurnal animals. For the most part, the smaller and nocturnal animals have 
been ignored in these environmental management plans. This trend is being 
reversed and ecologists and resource management specialists should be 
encouraged to include all biotic components in their management decisions. 
All biotic components should be recognized as a resource in their own right. 

One group of organisms that frequently are overlooked in this regard are 
the rodents. These small, secretive, and relatively unknown mammals 
seldom enter into consideration when resource management plans are 
drafted. This viewpoint is perhaps justified in some cases, when the affected 
populations are large or widespread locally and, therefore, are not subject to 
total elimination from the environment by increasing human activity. How- 
ever, when populations of any organism are few in number or limited geo- 
graphically, special consideration needs to be given to preservation of their 
habitat to prevent loss of what may be unique genetic, physiological, or 
environmental associations. 

The Guadalupe Mountain vole, Microtus mexicanus guadalupensis, may 
constitute one such rodent. It occurs typically in small, frequently isolated 
populations and, in some portions of its range (specifically that which 
includes the Guadalupe Mountains National Park) may be threatened by 
increasing human activity. 

The Guadalupe Mountain vole is an inhabitant of the Transition Zone of 
the scattered mountains of central New Mexico. Here it is found in the dry 
bunchgrass meadows between stands of yellow pine and fir, generally above 

395 



396 



WILHELM 




Fig. 1. Geographic range of the Guadalupe Mountain vole, Microtus mexicanus 
guadalupensis. The vole inhabits portions of the (1) Manzano; (2) Capitan-Sacra- 
mento; and (3) the Guadalupe Mountain ranges of New Mexico. 



2000 m but occasionally somewhat lower on northern exposures and in pro- 
tected canyons. It is more tolerant of xeric conditions than other species of 
microtines inhabitating this area (Findley and Jones 1962), and may be 
found at considerable distances from any permanent water source. 



STATUS OF MICROTUS 397 

The majority of the mammal distribution patterns in the southwestern 
United States may be explained as a result of an oscillating series of boreal 
expansions and contractions in the late Pleistocene (Findley 1969). These 
historical events have left populations of boreal mammals, including the 
Guadalupe Mountain vole, isolated on the scattered mountain ranges of the 
area, where they inhabit the remnants of the boreal forest at the higher eleva- 
tions. Because many of the mountain masses involved are rather small in 
total area, the boreal habitats are correspondingly small. Thus, because of 
this restricted habitat, the relict populations of boreal mammals also are fre- 
quently rather small, both in distributional area and in total numbers. The 
scattered nature of these southwestern mountain ranges makes it quite un- 
likely that recolonization of areas from which these boreal relicts might be 
extirpated would ever occur (Brown 1971). The effects of man's activities on 
these small, isolated populations and their habitat are therefore of concern if 
we are to preserve these associations for future generations. 

The geographic range of the Guadalupe Mountain vole encompasses the 
Manzano, Capitan, Sacramento, and Guadalupe mountain ranges (Fig. 1), 
and is separated from the other subspecies of Mexican voles by the Rio 
Grande Valley. These mountain ranges are rather restricted in area, and 
except for the Capitans and Sacramentos, are separated by areas of arid 
habitat unsuitable for microtines. 

Habitat suitable for the voles is not, of course, uniform within these 
mountain ranges, but is located primarily along natural drainage areas. This 
results in a pattern of more or less isolated populations of the vole within 
each mountain range, many of which are small in size, and in the Guadalupe 
Mountain vole being listed in the "status undetermined" category of the 
"Threatened Wildlife of the United States" published by the U.S. Depart- 
ment of the Interior. The occurrence of the Guadalupe Mountain vole at 
scattered localities within the Guadalupe Mountains National Park is there- 
fore of interest not only because the vole constitutes a part of the present 
fauna but because its continued existence within the park may be 
questionable. 

Within the park itself, the vole is apparently limited by the presence of 
bunchgrass meadows of creeping muhly (Muhlenbergia repens), needle- 
grass (Stipa tenuissima), and blue grama (Bouteloua gracilis). Suitable areas 
of these grasses are scattered over the park, primarily in some of the steeper 
canyons and on north-facing slopes. Populations of the Guadalupe 
Mountain vole have been found in Upper Dog Canyon, Lost Peak, The 
Bowl, Guadalupe Peak, and near Bush Mountain (Fig. 2). Interpretation of 
high altitude photographs of the park and field work of others have 
suggested that suitable habitats may also be found in other areas indicated in 
Fig. 2. These areas, which total approximately 650 acres, comprise less than 
1% of the total park area. Although these areas of extensive grass cover con- 
stitute the primary habitat for the voles, individuals have been trapped in 



398 



WILHELM 




Fig. 2. Verified and possible populations of the Mexican vole, Microtus mexicanus 
guadalupensis, in Guadalupe Mountains National Park. Verified populations are 
located at: (1) Upper Dog Canyon; (2) Lost Peak; (3) The Bowl; (4) Guadalupe Peak 
Campground; and (5) Bush Mountain. Possible habitat for the vole exists at (6), a 
stand of scattered aspen midway between The Bowl and Bush Mountain; (7) Upper 
McKittrick Canyon; (8) PX Flat; (9) The Cox Tank area; and ( 1 0) West Dog Canyon. 



Upper Dog Canyon on steep hillsides where there was a minimum of grass 
cover and on rocky hillsides dominated by agave {Agave lecheguilla), bear- 
grass (Nolina micrantha), and sotol (Dasylirion leiophyllum) with scattered 
bunchgrasses. In both cases however, extensive areas of grass were no more 
than 100 m away and these individuals were probably forced to these less 
desirable habitats by relatively high population densities in the more favor- 
able habitat. 

Previous studies have indicated that the Guadalupe Mountain vole may 
be found on the dry ridgetops, beneath shinnery oak where little grass cover 



STATUS OF MICROTUS 399 

was present (Bailey 1905, 1931). However, a somewhat more recent survey 
found voles only in grassy areas of the park (Davis 1940). An extensive 
survey of small mammals within Guadalupe Mountains National Park 
carried out during 1973-74 indicates that while the Guadalupe Mountain 
vole may be found in these dry, seemingly atypical microtine habitats, the 
meadow areas of the park support the bulk of the vole populations. These 
grassy areas apparently provide the primary refuges and in times of exces- 
sive population density, voles are capable of surviving for some time in the 
more xeric and unprotected habitats on hillsides and ridgetops. 

Because meadows or grassy areas are necessary for these voles, and 
because the same types of areas are ideal for campsites, picnics, and other 
human activity, a source of conflict is possible. In view of the rather fragile 
nature of the grasslands in the park, location of campsites or other high to 
moderate use areas within, or in proximity to, meadows could easily result in 
habitat degradation and consequent loss of valuable vole habitat. Over the 
northern portion of its range, the Guadalupe Mountain vole is found in more 
extensive habitat and is perhaps better able to withstand moderate amounts 
of temporary habitat loss. In the Guadalupe Mountains, however, any 
moderate loss of habitat, temporary or permanent, could put the vole in 
serious danger of being lost forever as a member of the park's fauna. 

The verified populations of the vole are not large. Although I have no 
population data to substantiate this statement, 2 years of field experience 
with the vole gives the impression that the density is maintained normally at 
a low level compared to other microtines. Trap-night data comparing the 
Guadalupe Mountain populations with populations to the north indicate 
that the densities in the park are only about one-half of those to the north. In 
addition, replacement of individuals lost to the population is slow. Because 
the extinction rate varies inversely with the population size (Brown 1971), 
this characteristic low density in habitats of restricted area makes these voles 
very susceptible to extinction of local populations. 

Recruitment of individuals to replace those lost to a population is largely a 
function of the reproductive strategy of the species. The small litter size of 
the Mexican vole has been pointed out by other investigators (Brown 1968; 
Choate and Jones 1970). Typical litters range from one to five young, with a 
mean of about 2.3. These small litters, however, are borne continuously 
throughout the year. This strategy is typical of a species operating near the 
carrying capacity of the environment. The low litter size requires only a 
minimal expenditure of energy for reproduction, with the reproductive 
stress distributed uniformly over the year. Because bearing and rearing of 
young constitutes a risk for the parent involved, this reproductive strategy 
probably represents a physiological adaptation to communities with intense 
intraspecies competition. Although this reproductive strategy may be 
well suited for the replacement of individuals lost to the population from the 
normal causes of mortality in a stable environment (such as the tropics), it is 



400 WILHELM 

less than adequate to rebuild rapidly populations that have suffered 
catastrophic losses. We evidently are dealing with a small mammal that has 
the reproductive strategy of a tropical montane rodent, but is living in a 
temperate montane habitat. Any large-scale population decrease caused by 
habitat loss may constitute a deficit from which the vole may never recover. 
For this reason, special consideration needs to be given to preservation of the 
few natural meadow areas of the park which provide the primary habitat for 
this small mammal. 

In this regard, the effects of fire on these natural areas of grassland is cause 
for concern. Any large-scale shift of faunal components of these meadows 
due to fire could place the Guadalupe Mountain vole in jeopardy. On the 
other hand, fire could conceivably enhance these habitats for the vole by 
shifting the faunal composition. Until our knowledge of the fire ecology of 
the meadow areas of the park allows prediction of the effects of fire on the 
floral components, particularly those plant species which most directly 
affect the Guadalupe Mountain vole, burning of these grasslands should be 
viewed with pessimism. 

Another factor to be considered in this regard is the effect of the amount of 
litter on the soil surface. There are some indications that colonization of 
burned habitat by voles is dependent upon the accumulation of a certain 
minimal quantity of litter (Cook 1959). This may mean that burned vole 
habitat could not successfully be recolonized for at least a year, or until at 
least one season's growth of vegetation had been converted into litter. 
Absence of this concealing litter would subject any voles that survived the 
fire to rather high predation losses, and in view of the low density and low 
reproductive value of the vole, would constitute a high probability of extinc- 
tion of that particular population. The effects of fire on the habitats 
favorable to the vole would seem to be of considerable importance to the 
proper management of the species within the Guadalupe Mountains 
National Park, and certainly merits further study. 

SUMMARY 

Assessment of the status of the Guadalupe Mountain vole within the park 
must take into consideration three factors — (1) the pattern of distribution 
(i.e., small isolated populations); (2) the low density that seems to be 
characteristic of the subspecies; and (3) the low litter size. Any factors 
impinging upon the voles or their habitat must be evaluated with these three 
factors in mind. 

The two problems that seem to be most important to the future of the 
Guadalupe Mountain vole are: (1) human impact on the grasslands 
inhabited by the vole (e.g., campsite locations, visitor traffic, hiking trails); 
and (2) the effects of fire on the preferred meadow habitats. 

Although the future of the Guadalupe Mountain vole within the park 
cannot be taken for granted, proper management procedures can maximize 



STATUS OF MICROTUS 401 

the chances that it will remain as a part of the unique fauna of the Guadalupe 
Mountains National Park. 

LITERATURE CITED 

Bailey, V. 1905. Biological Survey of Texas. N. Am. Fauna 25:1-222. 

1931. Mammals of New Mexico. N. Am. Fauna 53:1-412. 

Brown, J. H. 1971. Mammals on mountaintops: nonequilibrium insular bio- 
geography. Am. Nat. 10 5:467-478. 
Brown, L. N. 1968. Smallness of mean litter size in the Mexican vole. J. Mammal. 

49:159. 
Choate, J. R., and J. K. Jones. 1970. Additional notes on reproduction in the 

Mexican vole, Microtus mexicanus. Southwest. Nat. 14:356-358. 
Cooks, S. F., Jr. 1959. The effects of fire on a population of small rodents. Ecology 

40:102-108. 
Davis, W. B. 1970. Mammals of the Guadalupe Mountains of western Texas. Oc- 

cas. Pap. Mus. Zooi, Lactate Univ. 7:69-84. 
FlNDLEY, J. S. 1969. Biogeography of southwestern boreal and desert mammals. 

Pages 113-128 in J. K. Jones, ed. Contributions in Mammalogy, Univ. Kans. 

Mus. Nat. Hist. Misc. Publ. 51. 
Findley, J. S., and C. J. Jones. 1962. Distribution and variation of voles of the 

genus Microtus in New Mexico and adjacent areas. /. Mammal. 43:154-166. 



ACKNOWLEDGMENTS 

A portion of this research was carried out under National Park Service 
contract CX700040145 awarded to Dr. H. H. Genoways and Dr. R. J. 
Baker. I am also indebted to Dr. E. B. Fish for photointerpretation and to 
the National Park Service and its personnel for their assistance. 



Food Habits of Mule Deer on 
Foothills of Carlsbad Caverns 
National Park 



WALTER H. KITTAMS, Weiser, Idaho 

STANLEY L. EVANS, Carlsbad Senior High School, 
Carlsbad, New Mexico 

DERRICK C. COOKE, U.S.F.S., Santa Fe National Forest, 
Santa Fe, New Mexico 

The mule deer (Odocoileus hemionus) has been the major forage con- 
sumer on Carlsbad Caverns National Park since livestock use ended in the 
1940s. Warnings of range overuse by deer in the early 1960s raised numerous 
questions about deer-range relationships. Many of these questions could be 
answered only after the diet of the park deer was established. 

No studies of deer food habits had been conducted in the park. Data on 
areas comparable to the higher mountainous portion of the park were avail- 
able from Anderson et al. (1965). However, similar information useful in 
establishing the diet of deer on the lower foothills was meager. Therefore, 
this study of deer diet on the park foothills was initiated in 1967 as a major 
research project. 

Data collection ended in 197 1 . Circumstances precluded final analysis and 
report preparation before this senior author retired from his position of 
Research Biologist and Evans terminated as Biological Technician, in early 
1973. 



STUDY AREA 

The area covered by this study is the east end of Carlsbad Caverns 
National Park and adjacent area shown in Fig. 1 . It included about 6900 ha 
(17,000 acres) over two-thirds of which is in the park. 

Most of the park area is part of the limestone reef which forms the 
Guadalupe Mountains. The reef slopes gradually upward to the southwest 
and is dissected by the relatively shallow Walnut Canyon and deep Rattle- 
snake Canyon which drain to the east. Canyon sides are often very steep. 

403 



404 KITTAMS ET AL. 



• 'Upper Walnut Canyon 



Scoggins' Corner 



7 

Middle Walnut Canyon To Carlsbad 



Oak Spring ^Carlsbad Caverns 



s-» ^ 



k:^\ 




' ■ ' N Rattlesnake Canyon 




Guadalupe Escarpment 



Fig. 1. Study area for deer food habits research, Carlsbad Caverns National Park. 
Heavy solid represents surfaced roads, dashed line represents unsurfaced roads, 
and the widely spaced dotted line represents the boundary of the study area. 



Elevations of the area range from 1070 m (3500 ft) on the plain below the 
escarpment to 1520 m (5000 ft) on the upper ridges. The fine textured soils 
derived from limestone are generally shallow, seldom more than 15 cm deep 
on south facing slopes and ridges. Greatest soil depths are on canyon bot- 
toms and on the low plains. 

Climate is semi-arid continental with 78% of the annual rainfall (36 cm or 
14 in.) at the Caverns Station falling from May through October. Droughts 
are common. Recorded annual extremes have been 1 1 cm in 195 1 and 1 10 cm 
in 1941. 

Forage growth is mostly in early response to precipitation; shallow soils 
mixed with rocks and gravel result in low moisture retention. Precipitation 
records suggest that the 1966 growing season was generally good, perhaps a 
significant factor in the supply of cholla fruit (Opuntia imbricata) during the 
1967-68 period. Forage growth was limited severely in 1967 by a drought 
which began in October 1966 and extended through October 1967, with pre- 
cipitation normal or above only in June. Precipitation the following winter 
and summer was generally above normal except for June 1968. The 1968 
growth of forage, including cholla fruit, was good with forbs abundant in 
late summer. Precipitation in 1969 was almost consistently below normal 
until October. Growth of shrubs and production of cholla fruit were poor 
although there was a good crop of catclaw mimosa (Mimosa biuncifera) and 
redberry juniper fruits (Juniperus pinchotii) late in the season. 

Above-normal precipitation in the 1969-70 winter promoted spring 
growth, but 1970 spring and early summer moisture was down, and pro- 
duction of cholla fruit and other browse was only fair. Late summer pre- 



MULE DEER FOOD HABITS 405 

cipitation promoted a good supply of forbs. Heavy hail in early October 
1970 nearly eliminated succulent forage on the western side of the study area. 
Consistently below normal precipitation from November until late July 197 1 
caused severe drought in which shrubs suffered dieback (perhaps associated 
with subzero temperatures in January). A definite shortage of succulent 
forage prevailed in spring and early summer 1971, and forbs and half-shrubs 
produced only limited forage in late summer. 

A shrub-succulent community occupies the ridgetops. Dominant plants 
are sotol (Dasylirion leiophyllum), lechuguilla {Agave lecheguilla), oak 
{Quercus spp.), and redberry juniper. These are joined by shaggy mountain 
mahogany {Cercocarpus breviflorus), catclaw mimosa, and skeleton golden- 
eye ( Viguiera stenoloba) on the west, and ocotillo (Fouquieria splendens), 
prickly pear {Opuntia spp.), and Wright aloysia {Aloysia wrightii) on the 
east (Glass and Reisch 1972). Common grasses are three-awns {Aristida 
spp.), curlyleaf muhly {Muhlenbergia setifolid), gramas (Bouteloua spp.), 
and slim tridens ( Tridens muticus). 

The canyon walls are characterized by desert myrtlecroton (Bernardia 
obovata), oak, mountain mahogany, and various daleas {Dalea spp.). The 
canyon bottoms support a shortgrass-shrub community. Sotol, redberry 
juniper, lechuguilla, and cholla are common. Prickly pear, aloysia, and 
mariola parthenium {Parthenium incanum) enter the composition in the 
lower regions. Mescat acacia {Acacia constricta) dominates the gravelly 
sites. Creosote bush {Larrea divaricata) controls on the low, gravelly plains. 
Grasses at the low evaluations are black grama {Boutelouaeriopoda), hairy 
tridens ( Tridens pilosus), fluffgrass ( Tridens pulchellus) and three-awns. 

Deer were the only ungulates on the park range during the study. Cattle, 
sheep, and deer used the adjacent area. The deer population on the park 
foothills reached a relatively high level in 1966-67, dropped sharply by the 
next year, then gradually built to a moderately high level in 1970-71 before 
dropping again. 

METHODS 

The greatest amount of data came from direct observation of feeding deer, 
mostly in the morning but occasionally in the evening, between June 1967 
and December 1970. Most of these observations were along the main road 
between White's City and the Caverns Visitor Center, service roads in the 
vicinity of the park headquarters, and along the ridge road and canyon road 
that extend west to the head of Walnut Canyon. The usual procedure was to 
drive a car and watch for a feeding deer. After stopping the car, one observer 
would watch the deer's mouth with field glasses to try to establish the exact 
feeding site. The other observer would use a stop watch to record time, and 
he would watch for landmarks by which to direct his partner to the feeding 
site. It was important that the man without field glasses keep his position in 
the automobile so he could maintain a sighting and direct his partner to the 



406 KITTAMS ET AL. 

site after the deer had ceased feeding, or the men had reached observational 
limits. An observation was recorded only when browsing evidence was 
found on close examination. (This precluded erroneous observations 
possible when plants were close together, but it also caused us to discard 
some observations of forbs eaten in entirety.) Plant parts eaten and feeding 
time to the nearest tenth of a minute were recorded for each taxon. If field 
identification was not possible, a plant sample was taken for later identifica- 
tion. 

A separate entry was made for each continuous feeding (without a break 
of a tenth of a minute). Each feeding was termed an "observation" which in 
most cases was on one plant, if a shrub or large forb. But, if feeding was on 
small forbs, grass, or shrub clones, observation referred only to a con- 
tinuous feeding on the taxon involved. A maximum feeding time of 7 
minutes was established for record of individual observations. 

The original plan was to collect 50 paunches from deer killed in the park 
between November 1967 and April 1968. However, due to litigation, 
collection was terminated in December 1967 after 15 paunches were col- 
lected. Those deer were taken on representative portions of the entire study 
area, six in November and nine in December. 

Paunches also were salvaged from deer killed by automobiles in and near 
the park between June 1967 and December 1971. Of the 24 paunches col- 
lected by this method, 21 were from deer killed in the park on the main road 
between White's City and the Caverns Visitor Center, and 3 were from deer 
killed along U.S. Highway 62-180 extending along the base of the hills near 
White's City. 

In addition, paunches were collected from deer killed by hunters on lands 
adjacent to the park. Nineteen usable paunches were obtained between 1967 
and 1970. Of these, 18 were near the "Scoggins' Corner" area at the upper 
limit of the study area, and 1 from the base of the escarpment near the mouth 
of Rattlesnake Springs Canyon at the lower limit of the study area. All 19 of 
these deer were taken in November (New Mexico deer hunting season) 
during the 4-year period. The New Mexico Game and Fish Department 
approved taking of paunches from all deer killed by automobiles and those 
killed by hunters outside the park. 

The routine procedure for taking the paunch sample was to knead the 
paunch, then slit open the rumen and hand pick the sample to be 
representative of the rumen contents. Early in the study two quarts of the 
material were saved, but later the material was squeezed by hand to elimi- 
nate liquid and only one quart was saved. In a few cases the rumen con- 
tained insufficient material so it was supplemented from the reticulum. As 
soon as possible, the sample was taken to the laboratory and washed over a 
fine screen before preservation in formalin. 

All paunches were analyzed by the point -analysis method (Chamrad and 
Box 1964). However, we followed two changes in procedure from the ori- 



MULE DEER FOOD HABITS 407 

ginal method. The paunch material was washed on 6- and 3-mm screens. 
Material too coarse to pass the 6-mm screen was chopped by hand so that 
finally it all passed that screen (A. D. Chamrad, pers. comm. 1968). The 
sample used for analysis was taken from the material retained by the 3-mm 
screen. It was washed into a pan and thoroughly mixed before water was 
drawn off, leaving the plant parts barely suspended (Chamrad 1966). One 
hundred points, termed "hits," were read for each paunch sample. 

Only limited effort was given to tracking deer in snow for feeding record. 
Tracks were followed and "read" for feeding stops. For most tracking, 
records were made for each plant fed on — an "instance" — and for estimated 
volume of forage utilized on that plant. An estimated volume unit (EVU) 
was arbitrarily established as equivalent of one fruit of cholla or three twigs 
of skeleton goldeneye. The rare occurrence of snowcover, the extreme diffi- 
culty for human foot travel over the hidden rocks and between the armed 
plants, and reduced activity of deer all presented problems in this method. 
Samples obtained between March 1969 and Janury 1971 were of variable 
size — from 25 to 100 EVIFs and from 25 to 100 instances — due to experi- 
mentation and to limits of data on some tracking sites. All were near roads 
between White's City and the head of Walnut Canyon. 

Data were analyzed by sectors within the study area when possible, and by 
season of the year. Two seasons were recognized: growing season (GS) and 
nongrowing season (NGS). The GS was arbitrarily established as May 
through October. Since almost 80% of the average annual precipitation falls 
during these 6 months and most of the forage growth occurs then, this was a 
logical unit for consideration of forage availability. The NGS, from 
November through April, is generally too cold and often too dry for any 
appreciable forage growth. 

Data were additionally analyzed with regard to available forage, as judged 
from field observations during the several seasons and from rainfall records 
at the park. The bulk of forage consumed during the NGS is from the pre- 
vious GS. (Cholla fruits though may persist from a year before.) The 1968 
and 1970 GS's were rated good; thus the 1968-69 and 1970-71 NGS's were 
good. The 1967, 1969, and 1971 GS's were rated poor; thus the 1967-68, 
1969-70, and 1971-72 NGS's were poor. 

One paunch, from 19 September 1971, was assigned to good GS because 
forbs were rather abundant then in the area of the auto kill, and the deer 
paunch analysis yielded 74 hits on forbs and only 22 on shrubs. (This 
deviation from using a 1 97 1 paunch in the poor GS category did not conflict 
with treatment of observed feeding data because there were no observations 
after June 1971.) Another paunch, from an auto kill at White's City, was 
deleted from treatment because 61 of the 100 hits were on Chinese elm 
( Ulmus pumila), belvedere (Kochia scoparia), and popiar (Populus sp.), pre- 
sumed to be from the town area. 

For two snow-tracking sites where only feeding instances were recorded, 



408 KITTAMS ET AL. 

EVU's were calculated by use of EVU.instance ratios on other sites. 

The senior author directed the study and participated in all phases. Cooke 
jointly participated in collecting deer paunches and in feeding observations 
until August 1968, and he also analyzed most of the 1967 and 1968 paunches. 
Evans resolved many of the plant identification problems with assistance 
from Dr. Barton Warnock of Sul Ross State University, and Evans and 
Richard Young, recently at the University of New Mexico in Albuquerque, 
participated in tabulation and initial analysis of data. 

Plant nomenclature follows Correll and Johnston (1970). 

All of the foregoing research was done while the authors were employees 
of the National Park Service. Final analysis of data and preparation of this 
report were a contributed effort by Kittams and Evans; National Park Ser- 
vice support for some costs encountered is acknowledged. 

RESULTS AND DISCUSSION 

Data for establishing the diet of deer on the foothills came from 1083.3 
min. during 664 observations of deer feeding, from paunches of 57 deer, 
and from evidence of deer feeding during snow cover with volume of forage 
estimated and instances counted. A total of 102 taxa were recognized as 
being eaten by deer. 

Growing Seasons 

All data for the growing seasons are presented in Table 1 . Most (95%) of 
the observed feeding time during good GS's was in 1968, which was con- 
sistently good and thus should enhance representation of good forage 
seasons. Likewise, 86% of the observed feeding in poor GS's was during 
1967, late in an extended severe drought, and should give a reliable measure 
of poor GS's. Taxa eaten by deer during growing seasons were 36 shrubs, 31 
forbs, and 3 grasses. 

Observed feeding shows a greater relative use of shrubs during poor GS's 
and more use of forbs during good GS's. The numbers of shrub taxa repre- 
sented in the two types of season were nearly the same, although there were 
considerably more data for poor GS's. Several taxa were represented in only 
one category. In good GS's, Roemer acacia (Acacia roemeriana) leaves with 
some stems and fruits, and sotol stalks which are produced with good 
moisture, led in observed feeding time and number of observations. Silver 
dalea (Dalea argyraea) leaves with some stems, catclaw mimosa leaves with 
some fruit and stems, and mountain mahogany contributed. In poor GS's, 
Roemer acacia leaves with some stems dominated among taxa taken in 
feeding time and in observations. Catclaw mimosa was important, and 
lechuguilla stalks and redberry juniper fruits with leaves and stems con- 
tributed. It seems curious that lechuguilla stalks, regarded as a choice food, 
were more important in poor GS's. The lower relative showing of redberry 



MULE DEER FOOD HABITS 



409 



TABLE 1 . Records of deer feeding on foothills, Carlsbad Caverns area during growing seasons 
1967-71. All table entries are on a percentage basis. 





Observations of deer feeding in 


Analyses of deer paunches in 




good and poor growing seasons 


good and poor growing 
3 paunches 12 pa 


seasons 












unches 




Good 


Poor 


from 


good 
Fre- 


from 


poor 




222.5 


158 


399.2 


225 




Fre- 




Min- 


Obser- 


Min- 


Obser- 


300 


quency 


1200 


quency 


Food Item 


utes 


vations 


utes 


vations 


hits 


in 3 deer 


hits 


n 12 deer 








Shrubs 










Acacia 


















angustissima 














1 


17 


Acacia constrict a 


T 


1 


1 


3 


6 


33 


2 


50 


Acacia roemeriana 


23 


20 


39 


34 


36 


100 


30 


83 


Agave lecheguilla 


2 


2 


7 


3 


1 


67 


6 


33 


Aloysia wrightii 






T 


T 










A triplex canescens 


1 


1 


1 


T 










Bernardia obovata 


3 


4 


1 


2 










Brickellia 


















californica 


T 


1 


T 


T 










Celtis reticulata 














3 


33 


Cercocarpus 


















breviflorus 


7 


6 














Chilopsis linearis 


1 


1 














Cissus incisa 














T 


17 


Clematis 


















drummondii 










2 


33 






Dalea argyraea 


8 


9 


3 


7 






4 


50 


Dalea formosa 


2 


3 


2 


4 


2 


67 


1 


25 


Dasylirion 


















leiophyllum 


20 


9 


T 


T 










Echinocereus 


















stramineus 










3 


33 






Fallugia paradoxa 


2 


3 


1 


1 






8 


67 


Fendlera rupicola 














T 


8 


Fouquieria splendens 


T 


1 


T 


T 










Gymnosperma 


















glutinosum 














1 


17 


Juglans microcarpa 


1 


1 


4 


3 






5 


58 


Juniperus pinchotii 


T 


1 


7 


7 






3 


67 


Mimosa biuncifera 


8 


9 


18 


12 






3 


50 


Mimosa borealis 


T 


1 














Opuntia spp. 


T 


1 


3 


1 


16 


33 


2 


16 


Opuntia imbricata 






2 


2 






2 


17 


Parthenium 


















incanum 


3 


1 


T 


T 


1 


33 


T 


8 


Prosopis 


















glandulosa 






1 


1 






T 


8 


Ptelea trifoliata 






T 


1 










Quercus spp. 










1 


33 


2 


33 



410 



KITTAMS ET AL. 



TABLE 1. (continued) 





Observations of deer feeding in 


Analyses of deer paunches in 




good and poor growing 


; seasons 


good and poor growing 
3 paunches 12 pa 


seasons 












unches 




Good 


Poor 


from good 
Fre- 


frorr 


i poor 




222.5 


158 


399.2 


225 


Fre- 




Min- 


Obser- 


Min- 


Obser- 


300 


quency 


1200 


quency 


Food Item 


utes 


vations 


utes 


vations 


hits 


in 3 deer 


hits 


in 12 deer 


Rhus microphylla 






1 


2 


1 


33 


8 


42 


Rhus trilobata 


3 


4 


1 


3 






T 


8 


Ungnadia speciosa 










1 


33 


T 


8 


Xanthocephalum spp 


>. T 


1 














Yucca torreyi 














T 


8 


Unidentified 


















shrubs 










1 


33 


1 


33 


All shrubs 


85 


77 


92 


88 


71 




82 










Forbs 










Acalypha 


















lindheimeri 


T 


1 






T 


33 


1 


8 


Artemisia 


















ludoviciana 


T 


1 






1 


33 


4 


8 


Bahia pedata 


3 


3 






1 


33 






Chenop odium spp. 


5 


4 














Commelina erecta 






T 


T 










Croton pottsii 






T 


T 






T 


8 


Dyschoriste 


















decumbens 














T 


8 


Dyssodia spp. 














T 


17 


Eriogonum spp. 


T 


2 


1 


3 


T 


33 






Euphorbia spp. 










1 


33 


3 


33 


Galium sp. 














T 


8 


Gaura spp. 














1 


8 


Gilia rigidula 














T 


8 


Hoffmanseggia 


















densiflora 














1 


8 


Ibervillea 


















tenuisecta 


1 


1 


1 


T 










Ipomoea 


















lindheimeri 


1 


1 


T 


1 










Machaeranthera 


















scabrella 














T 


8 


Melampodium 


















leucanthum 


T 


1 














Monolepis 


















nuttaliana 


T 


1 














Paronychia jamesii 














T 


8 


Plantago spp. 














T 


8 


Portulaca sp. 


T 


1 


T 


T 










Rhynchosia texana 


1 


1 


3 


3 


T 


33 


1 


25 



MULE DEER FOOD HABITS 



411 



TABLE 1. (continued) 





Observations of deer feeding in 


Analyses of deer paunches in 




good and poor growing 


! seasons 


good 


and poor growing 


seasons 












3 paunches 


12 paunches 




Good 


Poor 


from good 
Fre- 


from 


poor 




222.5 


158 


399.2 


225 




Fre- 




Min- 


Obser- 


Min- 


Obser- 


300 


quency 


1200 


quency 


Food Item 


utes 


vations 


utes 


vations 


hits 


in 3 deer 


hits 


n 12 deer 


Salsola kali 














T 


8 


Sida procumbens 










4 


33 






Siphonoglossa 


















pilosella 


T 


1 






15 


33 


T 


17 


Sphaeralcea 


















angust (folia 










1 


67 






Stenandrium 


















barbatum 


1 


1 






2 


33 


2 


33 


Thamnosma texana 










1 


33 


T 


8 


Viguiera dentata 


T 


1 


T 


T 










Viguiera 


















longifolia 


1 


2 














Unidentified 


















forbs 


1 


4 


1 


1 


3 


100 


1 


42 


All forbs 


15 


23 


6 


10 


29 




16 










Grasses 










Bouteloua 


















curtipendula (grn) 


T 


1 














Sorghum halepense 


















(grn) 






1 


T 










Stipa sp. (grn) 






T 


T 










Unidentified grasses 


















(grn & dry) 






T 


1 






2 


58 


All grasses 


T 


1 


2 


2 






2 




Unidentified plant 


















material 














T 


33 



juniper in good GS's than in poor GS's probably reflects a low preference, 
since the leaves and stems are always abundantly available. 

The number of recognized forb taxa observed eaten in good GS's was 
twice that in poor GS's, for which we had more feeding records. Ranking 
taxa were goosefoot (Chenopodium spp.) and bluntscale bahia (Bahia 
pedata) in good GS's, and Texas snoutbean (Rhynchosia texana) in poor 
GS's. 

Direct observations of feeding indicate that deer spent over twice as long 
feeding on individual lechuguilla plants in the GS as in the NGS. Fruit stalks 



412 KITTAMSETAL. 

were mostly eaten in the GS. Probably the deer's preference for the rela- 
tively scarce tender stalks and greater volume of the one stalk on each plant 
account for the longer feeding time than on leaves during the NGS. (Deer 
were eating small plants in entirety during the NGS.) The only other shrubs 
with appreciable use in both GS and NGS were redberry juniper and silver 
dalea, and each had the same feeding time per plant for the two seasons. This 
is rather surprising in view of the presumed difference between foraging 
conditions in the two seasons. 

Most of the auto-killed deer paunches from GS's were from poor ones. 
The paunches from both good and poor GS's were well distributed by 
months and years but the sample is very small for good GS's. 

Relative amounts of forage classes found in paunches agree with the 
observed feeding, indicating more use of shrubs in poor GS's and more of 
forbs in good GS's. However, in paunches, shrubs are consistently less and 
forbs more than in the observed feeding. This may be due to bias with feed- 
ing observations caused by feeding time on forbs being so short, or plants so 
fully consumed that we could not obtain data. However, quicker digestion of 
forbs than of shrubs, and the chance of not hitting low-occurring items in 100 
hits per paunch would similarly seem to bias paunch analyses. 

Good GS paunches had only half the recognized shrub taxa found in poor 
GS paunches. Forb taxa were much less in good GS paunches, probably due 
in part to the small sample during good GS's. Roemer acacia leaves were the 
major good GS item in hits and frequency, followed by prickly pear fruits 
and some mescat acacia leaves and stems. (Most of the prickly pear was 
Opuntia engelmannii but it was not separated from other species.) Among 
forbs, only hairy tubetongue (Siphonoglossa pilosella) and spreading sida 
(Sida procumbens), both recorded in only the September 1971 deer, had an 
appreciable number of hits. Poor GS paunches had a major showing of 
Roemer acacia, and several other items of note: Apache plume (Fallugia 
paradoxa), littleleaf sumac (Rhus microphylla), lechuguilla leaves, and little 
walnut (Juglans microcarpa) leaves with some fruit. The low percentage of 
poor GS hits on redberry juniper, even though present in two-thirds of the 
paunches, is surprising. Forbs of note were Louisiana sagewort (Artemisia 
ludoviciana) and euphorbia (Euphorbia spp.). Grasses were very minor, 
showing only in the poor GS's. 

Differences between the data from observed feeding and from paunches 
may be due to the different locations represented. 

Nongrowing Seasons 

Data for the NGS's are treated in two categories, good and poor (Tables 2 
and 3). Seventy-five percent of observed feeding during good NGS's was in 
1968-69; and 90% of the observed feeding time during poor NGS's was in 
1967-68; thus, the two types of NGS's are well represented. 

Recognized taxa eaten by deer in NGS's were 39 shrubs, 32 forbs, and 2 of 
the grass and sedge category. Twenty-five shrubs and 15 forbs were common 



MULE DEER FOOD HABITS 



413 



t_ 


(U 


ed 


W) 


U 


K) 


c 




c« 


o 


p 


OJ 


„E 


a 




«J 


o 




,G 


C 




O 


c 


O) 


o 


£ 


M 


Mi 



,_, 


o 


O 


3 


<u 


rj 


T3 




Vh 


a 


O 


< 


V5 




T3 


CO 




C 


O 





o 


'./-. 


u 


73 


C* 


0) 
Vi 




SJJ 


CN 


C 






w 


£ 


PQ 


O 
ft 


< 




H 





— TH ^ 

in S a 

C n 7 

- o o 






& 3 S 





V. 






0) 


1^ 




J= 


o 




u 

c 

3 


o 


o 


a. 


M 


o 


CN 


ed 




'-* 




r^ 






sC 






Os 






"~ 


09 

QJ 


•a 



£ 



a 4 



^o *o ■* a\ 






H <N 



-3 o 
o o 

C 00 






© ,-S 



co r~~ o r- oo 



cn H co ^ H 



O O co 

lO lO CO 



CN r-i o 



O io 



H <N 



si 

so 







>> S 


u. 


45 O 


M 


O 4) 




o O 


b 


3 *<»■ 




C W) 




o- c 




3 M 








ed a> 












O V, 

o .♦: 

00 £> 




S o 




-h O 




"->— o 








< a 




as . 

si 

CO « 




5 "S 




* o 




+-. o 








< 00 









CN ^h CO 



O O co 
«0 IO CM 



CN CO f-H 



H H 



^ oo t 






§1 

| I 

£ 1 



.3 



^> bo ^l 



CO 


to 
,0 






■si 

ft. 




3 


5S 


53 


^2 





C 


.^ 

c 




5 


&a 

2 


-S 


_2 


<U 


3 




s 




•2 


•2 

■53 


<&, 


53 


^ 
^ 


.5 
•S 




~f 


g 


^3 


53 


?s 


8 




Q 






»a 




r o 


V 


"5 


"5 


^3 


c 


O 





Q Q Q kj 



414 



KITTAMS ET AL. 



•o — 3 <=* 
SB?"!- 

>> N 2 c5 



KJ 



.5 o\ 

•3 2 

>>o 



3 i> 



MS 

* O 
c ° 



4) *d 

o O 

e ao 

=5 M 



* 2 









O 

o 



cr c 



O V3 

o ."S 



05 

1> 


n 


JZ 


o 


o 


o 


c 


GO 


3 


!_ 


cs 


a> 


a 


Cm 


n- 


pj 



« O 
«m O 



H O 
<U b0 



o 


Vj 


© 


*J 


-r 


X 




t* 


>» 


a 


o 


Qi 


c 


"O 


=3 


«t 


cr 


_C 






O 


t» 


o 


*J 


"* 


Xi 


\C 


t/5 


oo 


-O 


*"* 


o 


o> 




uo 


c 


o 


5 




O 


P* 




i/j 


c 


»T) 


2 



oo O 



CS r- ~* 



H o 






co oo oo »/-> oo r~ i^i oo w 

oo w> r— <— i c*- oo 



OO \—tO\^D^ 



T* c* -h h -< 



io o o io »n 
t-» m iT) (N <n 



O io o 



«t O -H 



»-i i/-> t— cs '— i 



ro ^) (N t-» «— i 



51 


•2 



a 
"Bo 


1 


.5j 

a. 


53 


5j 

2 


53 


e 


5g 


,5j 


.53 


2 


a, 




5) 




~ 




,3 

"5 


"a 


S 


^3 


Ct, 


fe, 


o 


s 



2 

to -o 

C 53 



2 5 






S 
a 

.53 .53 .53 S 



a ^ 
a ? 

a. 



1 g 

V 05 ~3 i< 

cu a a bo 

a -a -a 



53 

•2 

5J 
5J 

s & 

5J 53 
^ 53 






MULE DEER FOOD HABITS 



415 



CO y© 

■« a S 2 

"-Is 



s ? 



<L> M 

i ? 

as 



'a 



- «3 



8* 



>» 4) 

O 0) 

B -d 



B tJ 



ON 



H - _h H 



H ^o 



H <N 



*0 io io 



H <^1 <N 




H ^ ^ 



8 s 
t -2 






5 

3 

1 






£ 2 o 



^ ^ <o u 



a, a 

p , V3 

- I 1 

S-> ^ P 

S4 -£, 



q ttj Cjq kj 



416 



KITTAMS ET AL. 



•o ss isa $£ 



3 5 



P» 










i-i 

o 
o 




,3 




CJ 


O- 




c 




0> 


3 


0) 


9 


? 


> 


v© 


53 





, ~ < 


^ 


z 







0) •_ 

-c o 

as 



« "2 
o o 



-c 2 

y 2 

c o. 

3 IH 

« 4> 






S> c w 



.s 





>> 0> 




O 4) 


o> 


C "d 




3^ 




o* c 








8 S 




VO X 




IH 




>> 0> 




O 0> 


o> 


e. -d 




3 ^ 




cr e 



8 £ 





■d 


43 


o 


u 


o 


c 


00 


3 


*-l 


rt 


4) 









^H 




I""- 




ON 






2 


6 


'-3 


^H 


0) 


o 


cm 


■*"• 


i_ 


00 


a> 


vO 


i3 


0"\ 




r^ 




sO 




ON 




»— < 



B 5 

'-*- o 



c -d 

3 "* 



© ."H 



\0 ri 

00 X 

-h 



H H 



t— c-~ r- i> 



<n H 



io mioOio *o io »o >o io 

<N <N <N U-> <N <N <N <N <N <N 



H«^H -< H H ^H 





•d 


<* 5 




o 




< 


o 

60 


w^ .S 










to S 

1— 1 ^ 



I s 



=?? "S cO 



3 'Bb 



•2 « £ 

•£> -52 O 

till 



C; ^ 



&« J 



2 S 



5« 



3 

ll 



'S3 §o 

S o 

.a 3 

c -2 



x 

0) su 

o -2 



t/5 rt ^3 k". 

5 "d S o 

« 2 ^ S 

ft. On QC CO 



•S c 
s § 



S § 
Si 

00 Co 



MULE DEER FOOD HABITS 417 



3 5 



M 2 O 

> ft" 
°l 



« 2 

* 2 

= u, 

si £ 

so a 



2 m 

•e o 

la 



0) "o 

3 o 

C 00 

P3 <U 

in e3 



8* 






j o a; 

tU 3 vO 



as vo vo o> 



-< H H ^ ro 



VO lO 



H ^ ^ 



E-. rH _ 



H H 



H -h 



.5 o\ 

■3 ~< 

a> O 

Si °° 
8 & 

ON 



-2 o 
<£ o 



8 * 



or c 



3 TT 

a* c 



vO </5 

00 X) 

- O 



° s 



<o .S 






~ H 



LJ 



^ H 



^ rt cn 



<N CO lo 



-a 

0) 

=3 

c 



til 
-J 

< 



a, "a 



II 



Co Co 



Si 

•Si aJ 
•5? '5 



« IS 



C 



• 5 O 03 



418 KITTAMSETAL. 

to both growing and nongrowing seasons. Three more shrubs and one more 
forb were noted by all methods in NGS's than in GS's. But direct feeding 
records were greater by one-third in GS's, while nearly 3 times as many 
paunches were examined in NGS's. Thus, the following comparisons to GS 
findings would be due in part to those differences: less shrub (19 vs. 26) and 
forb (12 vs. 17) taxa recorded in direct feeding observations; more shrub (16 
vs. 10) and forb (17 vs. 14) taxa found in paunch examinations. It is 
interesting to note further that snow tracking, done only in NGS's, revealed 
four shrubs and three forbs not represented in direct feeding observations 
nor in paunches from NGS's. 

The above results from uneven sampling preclude a conclusion as to rela- 
tive variety of shrubs and forbs in the diets of the two types of seasons. 
However, it does seem significant that all taxa represented in only one 
season — 11 shrubs and 16 forbs in GS, and 14 shrubs and 17 forbs in NGS — 
appear to be minor items of the diet, with the exception of skeleton golden- 
eye in the NGS. 

Observed feeding time and number of observations show greater rela- 
tive use of shrubs in poor NGS's and of forbs in good NGS's. Numbers of 
recognized shrub taxa were relatively few and nearly the same (13 and 15) in 
good NGS's and in poor NGS's, even though nearly twice as much data was 
collected in poor NGS's. Ten taxa were in only one category. Cholla fruits 
with a few stems was the leader in good NGS feeding time and observation, 
but several other species were important: catclaw mimosa leaves and 
fruit, redberry juniper leaves and stems, silver dalea leaves and stems with 
some fruits, and mariola parthenium leaves and stems with some fruits. In 
poor NGS's, observed feeding was mostly on redberry juniper leaves and 
stems, cholla fruits and stems, and lechuguilla leaves. (Of these, only cholla 
fruits are thought to be quality food for deer.) Additionally, only prickly 
pear stems with a few fruits and skeleton goldeneye leaves, dry flower heads 
and stems were of consequence. Feather dalea (Dalea formosa) and broom- 
weed (Xanthocephalum spp.) were of importance only in good NGS's, and 
skeleton goldeneye in only the poor NGS's. 

A considerable variety of forbs was taken in good NGS's, in contrast to the 
one taxon in poor NGS's with twice the feeding time and observations. 
Limited grass use was observed in poor NGS's only. 

Three of the paunches from auto kills were from the 1967-68 period and 
two from the 1968-69 period. The group of eight paunches from deer killed 
on roads showed heavy use of shrubs in poor NGS's, and appreciable use of 
forbs in good NGS's. A greater variety of shrubs was eaten in good NGS's. 
Most shrubs taken in poor NGS's were also taken in good NGS's. Walnut 
(mostly leaves), prickly pear fruits and cholla fruits led in good NGS's, but 
they made up only one-third of the hits. Other plants well represented in hits 
and frequency were Apache plume and broomweed. In poor NGS's, most of 
the hits accompanying high frequency were on redberry juniper leaves and 



MULE DEER FOOD HABITS 419 

stems, lechuguilla leaves, and prickly pear pads, with an appreciable show- 
ing of Roemer acacia leaves and a few stems, and some mariola parthenium. 
(Roemia acacia browse is available in the nongrowing season only during 
April when new leaves and succulent stems appear and in late fall when a few 
leaves remain green.) 

Hunter kills from the "Scoggins' Corner" area were mostly from upper 
elevations where oak is common and mountain mahogany is present on 
some sites. Since all samples were from the same month, they should indicate 
diet differences due to forage availability. The greater number of shrub taxa 
from poor NGS's (two more) than from good NGS's may be due to the fact 
that the sample size is twice as large. Likewise, the fact that few forbs were 
found in the good NGS paunches with only one more taxon than from poor 
NGS paunches may be due to the relatively small sample. The pattern of 
relatively more hits on shrubs in poor NGS's and of more hits on forbs in 
good NGS's is less prominent than with the autokills taken over the full 
nongrowing season. Likewise, there is considerable similarity in number of 
hits and frequency between types of nongrowing seasons on the major 
items — oak leaves with a few stems, Apache plume leaves and stems, and 
walnut leaves with a few stems — which made up about two-thirds of the hits 
in good and poor NGS paunches. Greater use of mountain mahogany in 
good NGS's could well be a reflection of availability, and greater use of red- 
berry juniper in poor NGS's may be due to lack of better food. 

One hunter-killed deer was grouped with the study kills to give representa- 
tion of the deer diet over the entire foothills in late fall 1967. The high pro- 
portion of shrub material taken in late fall 1967 agrees with other data for 
poor NGS's. Lechuguilla leaves and redberry juniper leaves and stems were 
leaders, both with rather high frequencies, but other taxa contributed 
appreciably: oak leaves, and cholla fruits with a few stems, and lesser 
amounts of Apache plume and Roemer acacia. 

Many of the shrubs represented in the foothills-wide deer were also repre- 
sented in the "Scoggins' Corner" area poor NGS deer, but amounts of taxa 
varied greatly There was much more lechuguilla and redberry juniper taken 
foothills-wide, and much more oak, cholla, Apache plume, and walnut taken 
at "Scoggins' Corner," which suggests that drought conditions were more 
severe on the lower foothills with regard to late fall forage. 

Data from snow tracking were divided by type of season: 225 EVU's in 
good NGS's and 264 in poor NGS's. Tracking records appear to be well dis- 
tributed over the area between White's City and the head of Walnut Canyon. 
Forage classes utilized were 94% shrubs and 6% forbs in good NGS's vs. 88% 
shrubs and 12% forbs in poor NGS's. Leading items in good NGS's were 
cholla stems with some fruits 22%, and broomweed 18%; followed by 
Apache plume 14%, redberry juniper 13%, and skeleton goldeneye 8%. In 
poor NGS's, oak led with 30%. Other items of 8 to 12% were redberry juni- 
per, cholla stems, skeleton goldeneye, broomweed, and the forb pepper- 



420 



KITTAMS ET AL. 











o> 




00 

c 




c 
o 


o 
o 

Cu 


SO 




o 




M 
Cm 


•a td s 




M) 




g 






o 




u 


<< 


CU 




o 




+_1 




Q 




ex 




3 










C 






<u 




13 










M 
1) 
CX 


| 

a 


'S3 ^ to 




03 




a. 


H 




-o 




3 


CU 






o 






£• 


lO 5 




o 






< 


<N * 




60 








•— » 




tH 








cu 
















<4— I 

c3 








0> 










o 


vO 
P OS 




i 






o 


CU > ^ 




X 






ex 


•S W © 




C/5 






< 


o f 

<U 




c 








Q 




> 
crj 




4) 




On 




ccj 




-5 




£ cu <* 

3 * oo 




"5 




S3 

3 




13 1 ^ 

u o - 




u 




a" 


■d 


r- > ^ 




^ 




•g 


8 


oo w 3 

cu ON 




0\ 

On 




a> 


00 

< 


C/5 

3 
E/2 


C/3 

<u 








J3t 

■a > °? 


■a 








^ W VO 




T3 
CU 












o 








l-» 




in 
















*CU 
c/5 








5 £ 




c 
o 

CU 

> 
o 









>& 




CM 


e 
o 
>« 

c 
in 
U 


o 

ex 

M 
CU 










< 


© is 




o 


CU 


-^ 




to S 




c 


[tj 


3 












C 








c 

CD 
O 


•a 


T3 


cu r^ 




3 
c/5 


cu 

a, 
en 


o 


1 

M 


CD jj ^ 
'53 >• cxT 




"ex 


c 


cu 

s 


J 


CU 

CM 


si 




o 


c/5 






c 


.33 






o 
















N 








gj 


tu 








3 






"5 


C3 


E 
cu 




cu 


=3 


•ia 




a 


< 


•o 






c/5 


8 




en 

W 


C 

o 

CO 


u. 




J 


CT3 

0j 






« 


CO 






< 








H 













m oo <^ 



•2 r 3 

II 

Cy ^: 

o ^ 
•2 



CU 



<u 



o 

$ n a 

O^ cj ^ 



a^ 



^ ^ ^ QQ 05 



Qq <o Q 



■I 

C3 C3 

Q Q 



2. « 



■a o 



& -2 

5£ 



o 

.5 .„ g 

•S o K 



ex -2 

ex 5 

C/5 -S 

"3 .« 



^ 5.- 

1 §-§ I 

S ex o ° 

s •= c S 3 S t Si 



- § g 

«■ «- Cy 



3 3 *- cu 

cx sx 



MULE DEER FOOD HABITS 421 





o 


c 


o 


o 


Cl. 


>, 


•h 


5 


4> 


u 


< 






3 




C 




13 






"8 


a> 


o 


D, 


60 


Q. 


In 


o 


■w 












w 



io 



*> > - 

■s w © 

_ O m 

—4 <-> 



T3 SO 

as *i ^ 

U D £ 
r- > • 

S w 3 



I > °p 

- W so 



•M "> -T 



o\ 00 o 



<u 


f» 


W5 * 


Oi 


£ 5 


"■" 


■a > 


00 


m H 






c 


>o 


n 


<N 


Hj 



-22 
•d 

§ 



p. 

t>5 Q V3 

^£ o ^ 
o sT ^ -S 

S -a ? 8 

sill 



11 



If 

09 .a ~ 

o, a o -S 






8 § 



5s i- y s r .«) n j? w c 



422 KITTAMS ET AL. 

weed (Lepidium virginicum). Oak was used only in poor NGS's, redberry 
juniper was the same in good and poor NGS's, and shrubs ranked lower and 
forbs higher in poor NGS's, suggesting that factors other than type of season 
were involved. Accordingly, data were treated by individual sites (Table 3). 
Observations from Middle Walnut Canyon were omitted because of the 
small sample (32 EVU's). 

Snow tracking samples from Lower Walnut Canyon showed nearly equal 
amounts of shrubs and forbs eaten in good and poor NGS's, but composi- 
tion differed. Apache plume leaves and stems, skeleton goldeneye leaves and 
stems, and cholla stems and fruits made up two-thirds of estimated volume 
in good NGS's; whereas, in poor NGS's, cholla stems were over one-third of 
the volume, and skeleton goldeneye, redberry juniper leaves and stems and 
the forb draba (Draba spp.) together made up one-third. Draba and other 
succulent mustards (Cruciferae) probably were available in March because 
of the above normal winter precipitation. 

At headquarters, a ridge site, nearly all volume was browse. Recorded 
EVU's in good NGS's showed broomweed leaves, stems, and fruits as a high 
leader, followed by cholla stems and fruits and juniper leaves and stems. Cal- 
culated EVU's from feeding instances near headquarters differed, having 
feather dalea leaves and stems a strong leader followed by lechuguilla leaves 
and some silver dalea leaves and stems. The differences between observed 
and calculated EVU composites suggest samples were too small for the 
different times or areas sampled. The poor NGS sample at headquarters had 
a different leader: oak leaves and stems, with appreciable amounts of 
skeleton goldeneye and broomweed leaves and stems. 

At Upper Walnut, all deer tracked had fed entirely on shrubs. Prickly pear 
stems and fruits and redberry juniper and Apache plume leaves and stems 
shared most of the good NGS volume. In poor NGS's, oak leaves and stems 
were more than half of the total, followed by redberry juniper leaves and 
stems, and less of tataiencho (Gymnospermaglutinosum) leaves and stems. 

Considering all three sites in both types of seasons, the most volume con- 
sistently eaten was of redberry juniper, followed by cholla, skeleton golden- 
eye, and broomweed. Additional study is needed to perfect the technique of 
estimating volume and to establish the effect, if any, of snow cover on the 
deer's feeding habits. Although snow cover usually lasts only a few days each 
winter on the foothills, tracking deer in it gives much more data for the hour 
of effort than does direct feeding observation. 

Generalized Diet 

A generalized diet of deer on the park foothills was developed from data 
presented here with consideration of sample size, time of year and location, 
and distribution and growth habits of plants. Shrubs comprise the bulk of 
the year-round diet. During the average growing season, Roemer acacia 
leaves with some stems is by far the leading item. Other shrubs of some 



MULE DEER FOOD HABITS 423 

importance are catclaw mimosa leaves with some fruits and stems; lechu- 
guilla fruit, stalks, and leaves; silver dalea leaves and stems; prickly pear 
fruits and pads; redberry juniper leaves, stems, and fruits; walnut leaves, 
fruits, and stems; Apache plume leaves, stems, and fruits; mescat acacia 
leaves and stems; feather dalea leaves and stems; and sotol fruit stalks when 
available. Top ranking forbs are tubetongue, shaggy stenandrium 
(Stenandrium barbatum), and sage wort. Roemer acacia is a less prominent 
leader during good growing seasons and sotol, prickly pear fruits, silver 
dalea, mescat acacia, catclaw mimosa, and mountain mahogany are also of 
some importance. During poor growing seasons, browse use is especially 
high. Roemer acacia is supplemented by catclaw mimosa, lechuguilla stalks, 
redberry juniper, walnut, Apache plume, silver dalea, and prickly pear pads. 

Shrubs comprise even more of the diet during the nongrowing season. The 
average diet then includes cholla fruit with stems and redberry juniper as co- 
leaders, followed by lechuguilla leaves, prickly pear pads and some fruits; 
catclaw mimosa leaves and some fruits; walnut; and silver dalea. At upper 
elevations oak leaves and Apache plume are important. Occasional use of 
forbs — mostly mustards with some tubetongue, wild buckwheat 
(Eriogonum spp.), and fleabane (Erigeron spp.) — occurs when moisture and 
temperature are especially favorable. 

In good nongrowing seasons, cholla fruit is the leader, followed by cat- 
claw mimosa, prickly pear fruits, silver dalea, redberry juniper, walnut, and 
Apache plume, with some forbs if winter weather is favorable. In poor 
nongrowing seasons, redberry juniper probably ranks first, closely followed 
by cholla fruits and stems, and lechuguilla with some oak and prickly pear 
pads. 

Variety of forbs in the diet, and of shrubs to a lesser extent, is greater 
during good growing and good nongrowing seasons. 

In an average year, Roemer acacia is taken from May to November, 
mostly May-July when leaves and twigs are succulent. For other important 
taxa, relative availability appears to be a determining factor in time of 
utilization: 

Mescat acacia August through November; most October-November 

Redberry juniper October through April; most November-February 

Cholla November through April; most December-March 

Lechuguilla December through April; most March-April 

Prickly pear January through March. 

This pattern of plant use suggests that the critical foraging period for deer 
on the park foothills usually extends from January through March. 

Comparative Studies 

The Texas Game and Fish Commission (Uzzell 1958) collected deer to 
study food habits in the Trans-Pecos region in 1956 and 1957, covering a 



424 KITTAMS ET AL. 

period of sustained drought. One of the study areas, the Sierra Diablo 
Range, about 120 km southwest of Carlsbad Caverns, has limestone forma- 
tion similar to that of the Guadalupes, and the elevation overlaps that of the 
upper part of our study area. Vegetation was juniper-oak-pinyon wood- 
land, but the paunch analyses indicate it was similar to that in our study area. 
In a collection of seven deer between September and November, the top food 
item in volume and frequency was lechuguilla; littleleaf sumac and oaks 
(Quercus undulata and Q. pungens) contributed, with some Ashe juniper 
(J. ashei). A collection of 17 deer in December to February had oak leaves 
first in volume and frequency, followed by Englemann prickly pear pads, 
Ashe juniper, lechuguilla, and cholla fruit. Allowing for the variation in 
species of juniper, and considering that most of our prickly pear used by 
deer was Engelmann and that the September-November collections ex- 
tended into our nongrowing season, we find comparability with the deer 
diet on the caverns foothills. The position of lechuguilla is reassuring, 
because some questions have been raised locally about our rating of this 
plant with its viciously pointed leaves. 

The New Mexico study of deer food habits between 1956 and 1960 
(Anderson et al. 1965) was conducted in the Guadalupe Mountains, mostly 
on the Lincoln National Forest, which extended within 10 km of the Cav- 
erns study area. It was done mostly on juniper-pinyon woodland at eleva- 
tions above the park foothills. The 93 deer paunches analyzed for that study 
showed wavyleaf oak (Q. undulata), juniper (/. monosperma and J. 
deppeana), mountain mahogany and yuccas ( Yucca spp.) as leaders. Forbs 
exceeded even the browse throughout a year of heavy precipitation. The 
importance of oak where present is borne out in our collections on the higher 
foothills, and there is no doubt that mountain mahogany on the park's 
mountain range is a major food for deer. Yuccas taken on the forest are very 
sparse in our study area. Certainly we had no evidence of such heavy use of 
forbs as was recorded there. In total, the results of that study appear to be 
more applicable to the mountains in the park than the foothills. 

In late spring 1969, a drought period, deer were found to be eating bark 
from madrone trees (Arbutus xalapensis) in McKittrick Canyon, Guada- 
lupe Mountains National Park, to such an extent as to change the aspect of 
the canyon. The Texas Parks and Wildlife Department, in cooperation with 
the National Park Service, collected four deer in the canyon to verify the use 
of madrone bark. Results are given in Table 4. The canyon bottom is an ever- 
green woodland containing madrone and a variety of shrubs. It is of special 
interest here because of its differences from the caverns study area, even 
though only 40 km away. The canyon is narrow and deep, and the floor at 
1 585 m (5200 ft) elevation has permanent water. The high use of oak suggests 
that the deer were hard pressed for forage and could have been stripping the 
madrone bark because of the shortage of quality foods. Certainly a number 
of the species found as low-occurring items are regarded as choice foods on 



MULE DEER FOOD HABITS 



425 



TABLE 4. Plants found in paunches of four deer from McKittrick Canyon, Guadalupe Moun- 
tains National Park, 16 June 1969. 







Percent of 


Percent 


Food item 


Plant parts eaten 


400 hits 


Frequency 




Shrubs 






Acacia roemeriana 


Stems 


T 


25 


Agave lecheguilla 


Leaves 


T 


25 


Arbutus xalapensis 


Bark mostly, few leaves 


7 


100 


Ceanothus greggii 


Leaves 


T 


25 


Celtis reticulata 


Leaves 


T 


25 


Cercocarpus breviflorus 


Leaves and stems 


1 


25 


Dalea argyraea 


Stems 


T 


25 


Fallugia paradoxa 


Leaves and stems 


1 


50 


Menodora longiflora 


Leaves 


T 


25 


Mimosa biuncifera 


Leaves and stems 


9 


100 


Parthenium incanum 


Leaves 


T 


25 


Prunus virens 


Leaves 


T 


25 


Quercus spp. 


Leaves and stems 


70 


100 


Rhus trilobata 


Leaves and stems 


3 


25 


Rhus virens 


Leaves and few stems 


2 


50 


Unidentified shrubs 




T 


25 


All shrubs 


Forbs 


96 




Acalypha lindheimeri 




1 


50 


Dalea nana 




T 


25 


Erigeron spp. 




T 


25 


Euphorbia spp. 




T 


50 


Machaer anther a blephariphylla 




T 


25 


Unidentified forbs 




1 


50 


All forbs 


Grasses 


3 




Gramineae 




1 


75 



the caverns range, but no information is at hand on the quantity that may 
have been available. Several of the taxa found are among those taken by the 
caverns foothills deer. Proportions, of course, may well be a reflection of the 
different vegetation. 



LITERATURE CITED 

Anderson, A. E., W. A. Snyder and G. W. Brown. 1965. Stomach content 
analyses related to condition in mule deer, Guadalupe Mountains, New Mexico. 
J. Wildl. Manage. 29(2):352-366. 

Chamrad, A. D., and T. W. Box. 1964. A point frame for sampling rumen con- 
tents. J. Wildl. Manage. 28(3):473-477. 

Chamrad, A. D. 1966. Winter and spring food habits of white-tailed deer on the 



426 KITTAMS ET AL. 

Welder Wildlife Refuge. M.S. Thesis, Texas Technological College, Lubbock, 
Texas. 

Correll, D. S. and M. C. Johnston. 1970. Manual of the Vascular Plants of 
Texas. Texas Research Foundation, Renner, Texas. 

Glass, M. R., and R. E. Reisch. 1972. Summary of range conditions. Inter- 
Agency Browse Analysis Survey, Carlsbad Caverns National Park. 

Uzzell, P. B. 1958. Deer food habits study. Trans-Pecos Game Management Sur- 
very. Job Completion Report, Project No. W-57-R-5. Texas Game and Fish 
Commission. 



Biomes of the Guadalupe 
Escarpment: Vegetation, Lizards, 
and Human Impact 



FREDERICK R. GEHLBACH, Baylor University, Waco, 
Texas 

When an area becomes a national park, its biotic features must be inter- 
preted to the public. The interpretive scheme must be simple — its elements 
easily recognized in the field, its synthetic features readily understood — and 
the scheme should be of such general applicability that a park visitor will 
become familiar with living landscapes throughout his travels. The biome 
concept fulfills these particulars; because biomes can be named for their 
vegetative structure, the functional significance of this is comprehensible 
(e.g. evergreen versus deciduous species), and biomes are repeatable units of 
landscape on a continental scale. Moreover, park visitors can identify the 
vegetative appearance of biomes without necessarily having to identify the 
plant species and speak of biomes in familiar language. 

I have defined plant formations in the Guadalupe Escarpment region and 
now wish to test the scheme against faunal patterns since biomes are groups 
of plants and animals whose ecologic niches overlap to a greater extent than 
expected by chance. What animals make the best indicators of biomes? 
Surely they must have spatial niches congruent with particular vegetation- 
types and be conspicuous to park visitors. I have not studied mammals, but 
certain rabbits and squirrels have potential, whereas nocturnal rodents do 
not, except perhaps through signs of their activities. Birds must be con- 
sidered, for they are primarily visual and diurnal like man. Among ter- 
restrial reptiles, snakes are avoided through fear, but lizards have qualifica- 
tions similar to birds and are even more meaningful to man because they too 
are earthbound. Are lizards useful and objective biome indicators? I shall 
investigate this possibility. 

Regardless of whether plant and lizard niches are coincident, human 
impact may disrupt the structure'and distribution of biomes. The features of 
secondary succession caused by man are as important as those of natural 
biomes, expecially on the Guadalupe Escarpment only recently released 
from livestock grazing. Therefore, I shall present data on plants and lizards 

427 



428 GEHLBACH 

in an area of Guadalupe Mountains National Park protected from grazing, 
alongside an area grazed simultaneously, and provide temporal informa- 
tion on plant and lizard succession along a pipeline construction scar in 
Carlsbad Caverns National Park. First, however, I offer a brief review of the 
vegetative basis of biome pattern in the Guadalupe ecosystem and cor- 
roborative evidence for part of this pattern in the spatial-behavioral niches 
of certain lizards. 

VEGETATION 

Five plant formations (hypothetical biomes) are delineated on the Guada- 
lupe Escarpment between U.S. Highway 62-180 and the crest and between 
Walnut and Pine Springs canyons (Fig. 1). Each is substantiated by quanti- 
tative, physiognomic distinctions (Gehlbach 1967). The borders between 
formations in Fig. 1 are approximate centers of major change in the living 
landscape, where plant life forms that determine a particular formation are 
least coincident. The most xeric sites, denoted by the Desert Formation 
(Biome)-class, are lowland flats and south or west-facing slopes, whereas the 
most mesic sites, indicated by the Woodland and Forest Formation (Biome)- 
classes, are east or north-facing slopes inside canyons, the drainageways, and 
the Escarpment-top exposures. 

Vegetation is taller and plant biomass increases in the xeric to mesic 
environmental gradient. Dominant shrubs are microphyllous, grow singly, 
and are widely spaced on flats and gravelly hills at the base of the 
Escarpment. These give way to clonal groups and multiple species clumps 
dominated by succulent and semisucculent species — an entirely different 
physiognomy — on rock outcrops in the lowlands and canyon slopes. I 
believe the Shrub Desert Formation is indicated by the microphyllous 
species, Larrea tridentata, Flourensia cernua, and Acacia constricta, for 
example. By contrast, the prevalence of clones and other clumps with succu- 
lent and semisucculent shrubs is indicative of the Succulent Desert Forma- 
tion. Agave lecheguilla and Dasylirion leiophyllum are chief indicators, but 
Viguiera stenoloba and Juniperus pinchotii are also important. The growth 
form index (deciduous/ evergreen + succulent species) of the Succulent 
Desert Formation is 0.35, whereas that of the Shrub Desert Formation is 
1.12. 

Similar physiognomic differences obtain between the Deciduous and 
Evergreen Woodland formations. The first is denoted by broadleaf 
deciduous species, the second by evergreens, both broad- and needleleaved 
species. Growth form indices are 3.07 and 0.37, respectively. A tree stratum, 
averaging less than 5 m in height with an open canopy, distinguishes the two 
woodlands from the shrub-dominated deserts and the Coniferous Forest 
Formation with its taller trees and closed canopy. Broadleaved deciduous 
trees grow primarily at springs and in streambeds at low elevations but 
become tne dominant growth form on stream terraces and in canyonheads 



BIOMES OF THE GUADALUPE ESCARPMENT 429 



ELEV. 
m. ft. 




Coniferous 
Nw Forest 


7000- 






•2000 


Evergreen Woodland 




6000- 






•1700 




Am 

Deciduous 
> Woodland 


5000- 


Pm S P \ 

o X. 

\\ Uo > 

\ 7 




•1400 






4000- 

■1100 


ci \ Succulent Desert \ o° 
us Shrub Desert --^ \ 


Flats Open Slopes Canyon Slopes 
SWEN SWEN 


Terraces 

Springs 

Streams 




xeric 


mesic 



Fig. 1. Plant formations of the Guadalupe Escarpment, New Mexico-Texas. 
Positions of lizards in the environmental gradient, determined by polar ordination of 
maximum IV's are indicated: Us, Uta stansburiana; Pc, Phrynosoma cornutum; Ci, 
Cnemidophorus inornatus